Introduction
I would describe myself as a fairly good tuner. Not because tuning ever came easy to me, I have just spent a lot of time working with the optimizer. I still have troubles rewriting complex SQL statements from time to time but I will admit that 20 years of experience makes it easier on me than it used to be.

One of the areas I am strong at is using the proper indexes to influence SQL statement access paths. We know that an index can influence the optimizer to change an access path from a tablescan to a table lookup using an index ROWID. But indexes can also influence the type of join used as well as the order of the tables being joined.

Let's take a look at the benefits virtual indexes have and we'll finish this blog by reviewing how indexes can affect simple access paths, join methods and join orders.

Virtual Indexes
I use virtual indexes in both 9i and 10G. I like them because I can quickly do "what if" scenarios. I can create the index and present it to the optimizer without waiting for the index to become populated and stored on disk. This feature allows me to reduce the amount of time I spend tuning, which is always a good thing.

I also use virtual indexes when I am training myself and others. Most production environments do have test counterparts, but there are times when I need to teach many different indexing concepts and waiting for indexes to build often becomes a detriment to the learning process.

9I OEM Virtual Index Wizard
The Virtual Index Wizard feature is provided in the Tuning Pack option of 9I OEM. Although there are a couple of ways to activate the wizard, let's take a look at the one that I most often use. Since we are going to be naviating through the SQL Analyze tool, we might as well review some of the other features it provides.

SQL Analyze
9I OEM provides the SQL Analyze tool to view access path information. SQL Analyze is only available if you have purchased the tuning pack option for 9I OEM. This screenshot shows me navigating through 9I OEM's menus to activate SQL Analyze.

When you activate SQL Analyze, the first step the utility performs is to display a listing of all of the top resource consuming SQL statements that are contained in the library cache. If the statement has been flushed from cache, you won't find it in this display.

The tool lists all of the databases on the left side of the panel that you worked with in the past. As you can see, I am a pretty active user of SQL Analyze. I can double-click on any statement in the top SQL listing to view its access path information. I can also use the menu system at the top of the screen to create a blank worksheet for new queries that I can enter manually. I found the statement I wanted from the listing and double-clicked on it to activate the SQL tuning panel.

The panel displays the SQL statement, and like our other utilities, allows us to step through the access path the query is taking. I can click on the SQL drop down menu at the top of the screen to ask Oracle to explain the statement using the various optimizer modes that are available.

SQL Analyze Tuning Tools
SQL Analyze provide us with tools that facilitate the SQL tuning process. The tools menu at the top of the screen allows me to choose from three different wizards. We'll review the Virtual Index Wizard in just a moment.

The Hint Wizard allows us to choose hints from a drop-down menu system, while the SQL Tuning Wizard activates an intelligent advisor that provides us with SQL tuning recommendations. Once you have run the SQL Tuning Wizard a few times, it's pretty easy to see that it was the precursor to some of the more advanced utilities that are now available in Oracle 10G.

Virtual Index Wizard
The Virtual Index Wizard allows me to create virtual indexes on tables that my query is accessing. The key word in that last sentence is "virtual". Oracle does not permanently create the index. The intent of the wizard is to allow users to determine the potential performance impact a new index would have on the statement being analyzed. It is then up to the user to permanently create the index.

If I click on the Virtual Index Wizard menu selection, 9I OEM responds by displaying the Virtual Index Wizard introduction panel. When I click on next, the wizard displays a panel that allows me to select the columns for my virtual index. Clicking next again displays a panel that allows me to attempt to force Oracle to use the new index or allow it to choose the new index on its own. My personal preference is to allow Oracle to choose the index on its own without providing any additional prodding.

Clicking Next again displays the projected cost improvement the new index will have on the statement being analyzed. I used a full table scan on a rather large table as an example. Based on the display, we can be pretty sure that the new index will have a positive affect on our statement's performance. The last panel also displays a button that allows us to view the before and after access paths.

The Virtual Index Wizard does a fairly good job of estimating the impact that new indexes will have on a SQL statement. It is an excellent tool to use when you are starting your tuning education as well as evaluating access paths in Oracle 9I databases.

I was surprised when I found that this feature was not contained in 10G Grid Control. I continue to firmly believe that using a tool to quickly build a virtual index, determine if the optimizer will use it and then display a prediction of the index's performance boost (or degradation) is very beneficial to the tuning process. I used the virtual index wizard dozens of times and will continue to use it when I tune 9I (and earlier) databases.

Virtual Indexes in Oracle10G
Although Grid Control doesn't provide the same graphical interface to create virtual indexes as it did in 9I OEM, all is not lost. We can revert to the tried-and-true SQL*PLUS command line to build virtual indexes using the "NOSEGMENT" option. The only feature it won't provide us is the forecast of the statement's performance.

The format of this blog page doesn't provide me with enough space to show you the output from my SQL*PLUS demo. Not without a whole lot of editing anyway. So, I decided that the best option would be to create a separate HTML page for the 10G virtual index demo output.

The demo will show you:

• How to create virtual indexes in 10G
• How to configure your session to allow the optimizer to choose a virtual index as an access path
• The affect the "_use_nosegment_indexes" parameter has on access path selection
• The error messages returned if you attempt to manipulate virtual indexes
• Information stored in the data dictionary pertaining to virtual indexes

Here's a link to the 10G SQL*PLUS Virtual Index demo!

Now that we understand how to create virtual indexes, lets' take a look at how any index can influence simple access paths, join methods and join orders.

Index Affects on Access Paths
Let's use a nested loop join as a quick and somewhat easy example. In a nested loop join, one of the tables in the join is known as the outer table while the other table (the table being probed) is known as the inner table. Oracle reads a row from the outer table and uses the columns in the join condition to probe the inner table.

Take a look at the SQL statement and graphical access path display below:

select a.employee_id, a.last_name, b.department_id, b.department_name,
c.street_address, c.postal_code, c.city, c.state_province
from hr.employees a, hr.departments b, hr.locations c
where a.department_id=b.department_id
and b.location_id=c.location_id
and a.employee_id = 174
order by a.last_name;

We are joining three tables together HR.EMPLOYEES, HR.DEPARTMENTS and HR.LOCATIONS. Notice that we have two different nested loop join operations. Oracle only joins two tables together at a time. It then creates an intermediate result set and uses that result set to join to the next table. It can use any type of join method available for any of the join operations.

The indexes we have available to us are:

• EMP_EMP_ID_PK - index on the EMPLOYEE_ID column of the HR.EMPLOYEES table
• DEPT_ID_PK - index on the DEPARTMENT_ID column of the HR.DEPARTMENTS table
• LOC_ID_PK - index on the LOCATION_ID of the HR. LOCATIONS table

The optimizer has chosen to join the HR.EMPLOYEES AND HR.DEPARTMENTS tables first using the nested loop join method. It is on this join condition - A.DEPARTMENT_ID = B.DEPARTMENT_ID. HR.EMPLOYEES is known as the outer table and HR.DEPARTMENTS is classified as the inner table of the join.

The optimizer looked at the local predicate "A.EMPLOYEE_ID=174" and checked to see if there was an index that it could use to improve data access performance. It found the EMP_EMP_ID_PK index that is defined on the EMPLOYEE_ID column. It will access the index first and retrieve the ROWID for the row that has an EMPLOYEE_ID of 174. It will use the ROWID to probe the HR.EMPLOYEES table.

As stated previously, the optimizer has chosen the nested loop join operation. It will access the HR.EMPLOYEES table using the EMP_EMPID_PK index, retrieve the value for DEPARTMENT_ID and use it to probe the HR.DEPARTMENTS table looking for a match (A.DEPARTMENT_ID = B.DEPARTMENT_ID). If it finds a DEPARTMENT_ID value in the HR.DEPARTMENTS table that matches, Oracle will use that row in future operations. If it does not match, Oracle will discard it.

Notice that the optimizer has chosen to use the DEPT_ID_PK index that is built on the HR.DEPARTMENTS' DEPARTMENT_ID column. Since we are probing this table using the values we found for DEPARTMENT_ID from the HR.EMPLOYEES table, Oracle will use the index on the HR.DEPARTMENTS' DEPARTMENT_ID column.

The B.LOCATION_ID = C.LOCATION_ID is the join condition that is used in the second join operation. The intermediate result set is used as the outer table and the HR.LOCATIONS table is used as the inner table. The same type of probe occurs. Since we are looking for matches using the LOCATION_ID contained in the intermediate result set, the optimizer will choose the LOC_ID_PK index.

Here's how indexes can influence the type of join used and the order of the tables being joined:

• If we didn't have indexes on either of the first two tables being joined, the optimizer would most likely have used a hash operation instead of a nested loop join to join the tables together. The optimizer knows that the hash join operation is best used for joins that return larger result sets. It could also have influenced the join order of the tables. In the example, the optimizer could have chosen to join the HR.DEPARTMENTS table to the HR.LOCATIONS table first. It would have then used the intermediate result set created to probe the HR.EMPLOYEES table.
• The more indexed columns we have on a table, the greater the chance that the optimizer will use it as the outer table of the nested loop join. It can filter out more rows before it sends it to the join operation. The earlier it can filter rows out, the better.
• If we have an index on both join columns (i.e. indexes on the DEPARTMENT_ID columns for both the HR.DEPARTMENTS and HR.EMPLOYEES tables), Oracle will use the DEPARTMENT_ID index on the inner table but not the outer. Why is that? Oracle doesn't know if it has a DEPARTMENT_ID VALUE in the inner table that matches a DEPARTMENT_ID value in the outer table unless it reads each and every row of the outer table to do the match. Think about it. It knows that since it has to read every row from the outer table to perform the match, the only indexes it can use are the indexes built on local predicates (i.e. "A.EMPLOYEE_ID=174").
• If we have an index on the join column for one table and not the other, the optimizer will often favor the table that has the index on the join column as the inner table of the join. It knows it can't use an index on the join column if it is the outer table but it can use the index if it is the inner.

This is just one quick example to show you the impact that indexes have on access path selection. The key to success is to continuously experiment with the optimizer. One of the features that can speed up your experimentation is to use virtual indexes. Remember, this feature uses an "undocumented" parameter in 10G.

I would like to thank super-DBA Mark Shore for re-igniting my interest in both 9I and 10G virtual indexes. Marks has a natural talent for this profession which I often wish I had.

Thanks for Reading,

Chris Foot
Director of Operations
Remote DBA Experts
Oracle Ace

The key to being a successful Oracle DBA is that you must be willing to accept the fact that with each new database release, you will have to adjust the way you administer your environments. In my profession, I work with many DBAs from companies across the globe. The most successful administrators I have met are the ones that intelligently leverage the new features that are available in the releases of the database they are administering.

Running Commands and Scripts Against Multiple Targets
Being involved in an IT audit, like death and taxes, is inevitable. If you work as a technician for any length of time, you will be working with an auditor and some point. Since we provide remote database services, our organization is audited numerous times per year. Sox audits, security audits, HIPPA audits - you name the audit and we have been involved in it. We have an office that is left open full-time for onsite auditors.

Auditors often ask you to run scripts on the operating systems and databases your unit is responsible for supporting. In the past, you would review the scripts and then distribute them to team members to run on the environments they were responsible for supporting. 10G Grid Control makes this task a whole lot simpler. Grid provides you with the capability of running host commands and scripts against multiple host targets at the same time. Virtually any command or script that can be executed on a single host can also be executed on multiple hosts using 10G Grid Control R2. Here's a blog that shows you how to use 10G Grid Control R2 to run operating system scripts against multiple targets.

10G Grid Control R2 also provides an easy-to-use panel that allows administrators to run SQL scripts against multiple database targets simultaneously. Just like its host command counterpart, all you need to do is code up a SQL statement, choose a set of database targets, run the SQL against the multiple targets and review the output. The benefits that this feature provides is only constrained by the DBAs imagination. Here's a link to a blog that will show you how to run scripts against multiple database targets.

Restore Points
How many times have database recoveries been performed because of incorrect changes made to database data? A user deletes or updates "one too many rows" by mistake and the result is a time-consuming and error prone process to restore and recover the database to a point-in-time before the error occurred.

Let's go back to my days as an Oracle instructor…. A common question in the Oracle DBA backup and recovery class was "why can't I just roll the database back to remove unwanted changes instead of restoring the database from a backup and applying the redo logs to roll forward to a point-in-time before the error occurred?" The question was invariably followed by "don't you think that would be easier?" The answer was always the same "yes, it would be easier, but the Oracle database doesn't have that capability."

That was before Flashback Database came along. Oracle's Flashback Database feature provided a new tool in the DBA's recovery toolbox. Flashback Database allows the DBA to "roll back" a table, set of tables or the entire database to a previous point-in-time. Administrators often used a timestamp during the execution of flashback operations.

10G R2 enhances Flashback Database by allowing administrators to associate a user-defined name with a point-in-time. The user-defined name, called a restore point, can be used in place of a SCN or timestamp when executing a FLASHBACK TABLE or FLASHBACK DATABASE statement.

A SQL statement is used to create the restore point which means it can be embedded in application programs as we as executed on an as-needed basis in SQL*PLUS. The example below shows the creation of a restore point:

CREATE RESTORE POINT batch_daily_cycle_complete;

Now if I execute the following statement:

UPDATE cti.employee_salary_table SET
salary = 200000 WHERE empl_lname = 'FOOT';

My boss can execute this statement to correct the 'mistake' (although I prefer to call it a fully justifiable pay adjustment):

FLASHBACK TABLE cti.employee_salary_table TO batch_daily_cycle_complete

10GR2 also provides guaranteed restore points which ensure that the database can be flashed back to the point-in-time they were created. Guaranteed restore points save disk space because only the flashback logs required to meet the guaranteed restore point need to be retained by the database. The statement below creates a guaranteed restore point:

CREATE RESTORE POINT batch_daily_cycle_complete GUARANTEE FLASHBACK DATABASE;

Automatic Undo Retention
In Oracle9i, administrators had their choice of continuing to manage rollback segments on their own (manual undo management) or configuring the database to manage its own before image data (automatic undo management). Oracle refers to system managed before image segments as undo segments.

You didn't have to be an Oracle expert to know that manual rollback segments were "somewhat troublesome." Out of space conditions, contention, poor performance and the perennial favorite "snap shot too old" errors had been plaguing Oracle database administrators for over a decade. Oracle finally decided that the database could probably do a better job of managing before images of data than we could.

But implementing automatic undo retention didn't necessarily guarantee users a trouble free undo environment. There really is only one parameter that administrators can tweak in a system that uses automatic undo. The UNDO_RETENTION parameter specifies the amount of time in seconds that Oracle attempts to keep undo data available. Setting this parameter to the appropriate value could be described as more of an art than a science.

Set it too low and you are wasting disk space. In addition, you aren't taking advantage of being able to flashback your data to as far back as the disk space allocated to the undo tablespace allows. Set it too high and you are in danger of running out of freespace in the undo tablespace.

10G R2 comes to the rescue! The database now collects undo usage statistics, identifies the amount of disk space allocated to the undo tablespace and uses that information to tune the undo retention period to provide maximum undo data retention. Administrators can determine the current retention time period by querying the TUNED_UNDORETENTION column of the V$UNDOSTAT view.

RMAN Automatically Creates Temporary Datafiles During Restore
Ever perform a restore and forget to recreate the temporary tablespace's data files? I have. The database comes on-line but the first time a user's sort overflows from memory to disk, an error code is returned. Recreating a database's temporary files is one of those activities that are often overlooked. 10G R2 RMAN will automatically recreate temporary datafiles belonging to locally managed temporary tablespaces during the recovery operation.

Enterprise Manager Statistics Management Support
A new link titled 'Manage Optimizer' statistics has been added to the database administration home page in Enterprise Manager. Clicking on the link displays the new Manage Optimizer Statistics home page. The new home page provides wizards and templates that facilitate and automate optimizer statistics management. In 10GR2 administrators are able to use Enterprise Manager to gather, restore, delete, lock and unlock statistics

Why would you want to lock and unlock statistics? Let me describe one real-life example from a recent project. We have a warehouse system that contains an extremely volatile table. Describing it as volatile is like stating the Titanic sprung a small leak. Hundreds of thousands of rows are added and removed around the clock.

The level of activity is based on the particular business process taking place. At one particular time, the table can contain hundreds of thousands of rows and at other times it can contain dozens. OK, so when do you generate statistics? If you run DBMS_STATS at the same time each day, in most cases you would think you would get a predictable set of statistics generated. Not so in this table's case, sometimes you get hundreds of rows and some times you get hundreds of thousands.

If you are unlucky and generate statistics on the table when it has hundreds of rows, access paths suffer when the table grows to hundreds of thousands. So, we ran DBMS_STATS when we knew the table contained numerous rows and never generated statistics on it again. Problem is that we can no longer easily use the SCHEMA option for DBMS_STATS. Freezing statistics on this table will allow us to use the SCHEMA option and not generate statistics for selected tables.

DML Error Logging
What I like about SQL*LOADER is its ability to continue processing through load errors. If the record being loaded is rejected because of a unique constraint violation or discarded because it does not meet some user-defined criteria, SQL*LOADER places the record into a discard or reject file and keeps on running until it has reached a user-specified maximum number of rejects or discards.

The loader's log file will show how many records were loaded, rejected or discarded. I can look at the messages and review the discard or reject files, fix the problem and attempt to reload them again.

In 10G R2, this same type of processing has been applied to bulk DML operations. Users are able to specify whether they want to log errors or abort the entire DML statement, set the amount of detail information logged and the maximum error threshold. So instead of the entire DML statement blowing up and rolling back, Oracle will log the errors to the error table and continue processing until it exceeds the maximum number of errors. Just like SQL*Loader!

Oracle 10G R2's DBMS_ERRLOG package is used to create the error log output table and link it to the table being updated. The package's specification is provided below:

DBMS_ERRLOG.CREATE_ERROR_LOG (
dml_table_name IN VARCHAR2,
err_log_table_name IN VARCHAR2 := NULL,
err_log_table_owner IN VARCHAR2 := NULL,
err_log_table_space IN VARCHAR2 := NULL,
skip_unsupported IN BOOLEAN := FALSE);

Most of the columns are pretty self explanatory: table name being updated, error log table name, owner of error log table and the error log table's tablespace. If the SKIP_UNSIPPORTED is set to TRUE, column types that are not supported by error logging will be skipped over and not added to the error logging table. If it is set to FALSE, an unsupported column type will cause the procedure to fail.

Here's a quick example:

Let's create or table that will be updated:
SQL> r
1 CREATE TABLE foot.emp_table
2 (empno number(4), ename varchar2(10), job varchar2(8))
3* TABLESPACE users;

Table created.

Add a primary key:
SQL> ALTER TABLE foot. emp_table ADD PRIMARY KEY(empno)
2 USING INDEX TABLESPACE users;

Table altered.;

Load some rows:

SQL> INSERT INTO foot.emp_table VALUES
2 (7499, 'ALLEN', 'SALESMAN');

1 row created.

SQL> INSERT INTO foot.emp_table VALUES
2 (7521, 'WARD', 'SALESMAN');

1 row created.

SQL> INSERT INTO foot.emp_table VALUES
2 (7566, 'JONES', 'MANAGER');

Let's cause a unique constraint violation:

SQL> insert into foot.emp_table select * from foot.emp_table;
insert into foot.emp_table select * from foot.emp_table
*
ERROR at line 1:
ORA-00001: unique constraint (FOOT.SYS_C009069) violated

Create the 10G R2 error logging table:

exec dbms_errlog.CREATE_ERROR_LOG ('emp_table','dml_errors_emp_table')

Let's create a table that we can update and change one row to see if we can get one row to load and 2 to fail and be placed into the DML_ERRORS_EMP_TABLE:

SQL> create table foot.emp_load_table tablespace users as select * from foot.emp_table
Table created.

Update one row to change the EMPNO column's value to avoid the primary key violation:

SQL> update foot.emp_load_table set empno=123 where empno=7499;
1 row updated.

Rerun the statement specifying the new 10G R2 error logging syntax. Use our new load input table so that our one changed row will be loaded and two will be rejected and placed into the DML_ERRORS_EMP_TABLE:

SQL> insert into foot.emp_table select * from foot.emp_table
LOG ERRORS INTO dml_errors_emp_table('test_load_20050718')
REJECT LIMIT UNLIMITED;

1 row created.

OK, we loaded one. What happened to our other two? Let's see what our DML_ERRORS_EMP_TABLE contains.

First, let's describe the DML_ERRORS_EMP_TABLE:

SQL> DESC foot.dml_errors_emp_table

ORA_ERR_NUMBER$ NUMBER
ORA_ERR_MESG$ VARCHAR2(2000)
ORA_ERR_ROWID$ ROWID
ORA_ERR_OPTYP$ VARCHAR2(2)
ORA_ERR_TAG$ VARCHAR2(2000)
EMPNO VARCHAR2(4000)
ENAME VARCHAR2(4000)
JOB VARCHAR2(4000)

Our error logging table contains an incrementing error counter, the error message, ROWID, error type, tag (contains our users specified name from above -'test_load_20050718') and the three columns of our table.

Let's select from the table. I have truncated the ORA_ERR_MESG$ table for readability sake:

SQL> SELECT ora_err_number$, ora_error_mesg$, emp_no FROM foot.dml_errors_emp_table;

ORA_ERR_NUMBER$ ORA_ERR_MESG$ EMP_NO
--------------- -------------------------------------------------- ------
1 ORA-00001: unique constraint (FOOT.SYS_C009069) violated…..7521
2 ORA-00001: unique constraint (FOOT.SYS_C009069) violated….7566

We loaded one row and rejected two rows due to primary key constraint violations.

Conclusion
I hope you enjoyed this series on 10G features.

Favorite DBA Conversation #1
Any discussion that ends in "I hope you're not too busy because we need it now." I really dislike the conversation that ends in "we needed it yesterday." If you needed it yesterday you should have stopped in to discuss this request a week before yesterday. We all know these types of requests because we have all gotten them. One of your customers pops in and states:

"We need a complete refresh of XYZ database from production. Oh, and its an ERP application database and we don't have enough disk space in test to hold everything so you'll need to figure out how to reduce its disk size requirements AND make sure the application still works.

I know the box only has 1 CPU and 250 MEG of memory, but can you make sure we still get good performance? We're testing some application coding changes and we need to get performance benchmarks. We'll compare the stats from the test box to the ones we get from our 22 CPU production box. It may not be an "apples to apples" but I think the boxes are close enough to compare the performance stats.

I know its Friday afternoon at 3:00 and even though we knew about this refresh 14 weeks ago, we didn't think we needed to tell you until 1 day before we wanted it. So, we would like to have the system ready by Monday."

Over the last 20 years I have received dozens of these last minute requests. Requests that have made me cancel plans, postpone trips and generate numerous apologies to friends and family. The work never bothered me, but the timing of the requests sometimes did. It also reduced the quality of service I provided to my other customers. When you have DBAs jumping from one non-planned request to another, it affects their ability to work on planned requests as well as perform proactive administrative duties.

Fix #1: You minimize these requests by creating, advertising and enforcing DBA work request SLA agreements. Agreements that clearly state lead times per type of request. The more widespread the problem is at your organization, the more advertising you must do. It really becomes a sales pitch on the importance of a DBA unit being able to forecast workloads. The unit can't provide quality service to their customers if they are being affected by work requests that could have been planned for - but weren't.

The mantra should be "all of our customers suffer when we receive un-planned work requests." I'm not afraid of telling application manager A that she can't get her planned requests done because application manager B just gave us 32 emergency requests. Kind of like putting them both in a jar and shaking it up a bit.

Fix #2: 10G Grid Control allows you to clone both the Oracle Home software installation as well as the Oracle database. Grid will automatically adjust the specific home properties (IP address, host name, listener settings) during the cloning process. 10G DBCA templates allow you to clone a new database from existing databases. In addition, DBCA allows you to clone just the source database file structures or the structures and the data.

You can also use some of the 10G Data Pump features to tailor the new database to your user's specifications. 10G Data Pump provides numerous options that:

• Specifiy what objects you want to transfer from one database to another
• Transfer a subset of data from those objects using various filtering options
• Remap objects during the migration process. You can remap schema objects from one user to another, from one tablespace to another

You already know I'm a huge fan of 10G Data Pump. You need to leverage the features this tool provides to customize schema and data transfers from one database to another. You shouldn't be writing scripts when you don't need to. The only way you will realize the full potential of 10G Data Pump is to READ THE UTILITIES manual and see how powerful 10G Data Pump really is. It will make your life easier.

Favorite DBA Conversation #2
You come in to a meeting and you're confronted by an application manager who tells you that you need to tune your database because his ace developer "Sir Lawrence of LongSQL" can't get his 32 page SQL statement to achieve sub-second response times.

You know the SQL statement that I'm talking about. It's the one that contains 15 inner and outer joins, 12 subselects, and…….(feel free to add any other type of complex SQL Syntax here).

Note from database administrators to application developers:
Just because you can code a 32 page SQL statement, it doesn't mean you should. We all understand how powerful the SQL language is, but do you really need to try and use every feature available in a single statement? There's a difference between leveraging the power of the SQL language and attempting to perform too much work in a single SQL statement. Generating long, complex SQL statements that no one can debug does NOT classify you as a supercoder. It does classify you as something else, but I can't say that publicly.

Favorite DBA Conversation #3
Favorite DBA Conversation #3 combines components of conversations #1 and #2. It occurs when you are approached by a customer who says they have developed this "small" application that goes live in two weeks and it isn't performing "as well as we think it should." This means that you have to review the SQL from 23 new application programs in those two weeks.

Fix #1: Reinforce the need for the DBA team to be involved in all application development projects. This can be done by creating a series of application design review meetings. A dedicated an entire blog on this subject.

Fix#2: You will need to start a top down performance analysis project that focuses on tuning the most poorly performing queries first. If you don't have access to the 10G Grid Control sissy GUI tools, you can use the V$ views to retrieve the top resource consuming statements. You can use the V$SQLAREA view for Oracle9I and below. 10G offers the V$SQLSTATS view that provides some distinct advantages over its previous counterparts.

10G V$SQLSTATS Performance View
Before we discuss the new V$SQLSTATS view, let me deviate for a minute to provide you with some quick SQL tuning advice. V$SQLAREA is one of my favorite SQL tuning views. Until V$SQLSTATS came along…

If I want to perform a traditional "top down" tuning approach and tune the highest resource consuming SQL, I'll use the statements below to identify the top resource consuming queries.

The following query identifies the SQL responsible for the most disk reads:

SELECT disk_reads, executions, disk_reads/executions, address, sql_text FROM v$sqlarea WHERE disk_reads > 5000 ORDER BY disk_reads;

The following query identifies the SQL responsible for the most buffer hits:

SELECT buffer_gets, executions, buffer_gets/executions, address, sql_text FROM v$sqlarea WHERE buffer_gets > 10000 ORDER BY buffer_gets;

You can create a more readable report in SQLPLUS by inserting report breaks between the output lines. To generate the report breaks in SQLPLUS, issue the following statement before running the query:

BREAK ON disk_reads SKIP 2 --- for the disk read report and
BREAK ON buffer_gets SKIP 2 --- for the buffer get report

The first query returns SQL statements responsible for generating disk reads greater than 5,000 while the second query returns SQL statements responsible for generating buffer reads greater than 10,000. I used these numbers just as an example but I sometimes use them as a starting point. I'll then adjust them up or down accordingly, based on their output. The numbers also depend on the system I'm reviewing. I'll use different numbers for OLTP environments than I would for data warehouses.

You'll notice that I divide the number of disk and buffer reads by the number of statement executions. If a statement is generating 1,000,000 disk reads but is executed 500,000 times, it probably doesn't need tuning.

Heavy disk reads per statement execution usually means a lack of proper indexing, poor selection criteria, etc.. Heavy buffer reads sometimes means the exact opposite - indexes are being used when they shouldn't be.

But I'm personally most interested in workload, that's why I most often use the buffer cache hits in my initial queries.

But the SQLTEXT column in V$SQLAREA does not provide the entire text of the SQL statement. That's why I include the address column in the report. I can use that value to dump the entire SQL statement from V$SQLTEXT using the statement below (where xxxxxxxx is the value in the address column from the V$SQLAREA reports above):

select sql_text from v$sqltext where address = 'xxxxxxxxx' order by piece;

Oracle 10G R2 provides a new view called V$SQLSTATS that contains a combination of columns that appear in V$SQL and V$SQLAREA. The benefits that V$SQLSTATS provides are as follows:

• Since V$SQLSTATS contains the entire text of the SQL statement AND its associated performance statistics, we are no longer required to access both the V$SQLTEXT and V$SQLAREA to obtain the information we need.
• Oracle states that V$SQLSTATS is faster and more scalable.

The data in V$SQLAREA has a tendency to get its contents flushed out just when you need to get additional information from it. The V$SQLSTATS view provides users with a longer access window. That's one of the key benefits to this view.

Next Up
The series continues.

Ease of Installation
I remember the good old days. When I began my career, the Oracle installers were just screen painters that forced you to use the TAB key to navigate back and forth between selections and screens.

Installations used to take hours and experiencing problem after problem was commonplace. Successfully installing Oracle became a rite of passage as a DBA. Four hours after you started the installation (and fighting through errors, warnings and placing a half dozen CDs in the reader), you would finish up and think to yourself "I am an installation GENIUS!". You next thought would be "I hope I don't have to go through this again for a while."

With each new release, the installer became easier to use, installation speeds increased and the number of installation problems decreased. In a few years, I'll be telling new DBAs "You don't know how good you have it. I remember when installations used to take three days... We were real DBAs back then. We even had to log in at NIGHT TIME and take the data OFFLINE to do work." Darn kids don't know how good they have it nowadays….

We do DOZENS of Oracle installations a month here at Remote DBA Experts. One of the benefits of working here is the sheer thrill of working on virtually every operating system that Oracle makes a product on. Name the O/S and hardware combination and we can truthfully say "been there and installed an Oracle product on it."

When our first set of 10G CDs arrived a couple of years ago, the DBA opened up the package and said "Hey, where's the rest of them? I think they forgot to send me all of the CDs I need to do the install." We checked MetaLink and found that Oracle10G only needs a couple of CDs. One main CD and another called "The Companion CD" which really should be called "The Companion CD that is required for all installs" since it pretty much has to be installed when you install Oracle.

The team here really likes the fact that the installer now performs pre-installation requirement checks before laying down the software. The installer itself now determines if the O/S environment is configured to correctly run the Oracle software. Having the installer verify that you have performed "your pre-installation" duties is very beneficial.

10G Data Pump
I'm a big fan of 10G Data Pump. Once you become accustomed to the new environment, it's a pretty good utility. There is a learning curve - like the first time you try to restart a Data Pump job by just rerunning the script again. You'll quickly find out that Data Pump has a command for that. At least it provides you with the "Master table exists" error message to point you to your mistake.

Here's a quick list of some of the features I like:

• The utility IS faster than Export and Import. Maybe not as much faster as Oracle says it is - but it is faster. I have personally seen Exports and Imports run for hours - and days.
  
  
• I like being able to use parallelism to improve performance.
  
  
• You can stop and restart jobs. How great is that? The master table knows where you left off so you don't have to spend all of that time trying to code a new Import statement that won't duplicate rows in tables that were successfully loaded. This is really helpful.
  
  
• How many times did you have to "pick a number, any number" when you had to determine how much space you would need for your Export Dump file? Data Pump provides an option that estimates the output dump file size that would be created by a Data Pump Export.
  
  
• One of my favorite features is the network transfer option that allows administrators to attach to any source database that can be accessed from the network, transfer the data and write it to the target database. No dumpfile required! Administrators use the NETWORK_LINK parameter to specify a pre-created database link that points to the source database.
  
  
• You can use the Data Pump Flashback feature to go back in time and export data how it looked in the past. This helps when you get a visit from one of your developers stating "I think I might have updated 47 production tables by mistake, but I won't know until I see the data." Use the flashback feature during the Data Pump Export and use the Data Pump Import utility to pump the data into a new schema for the developer to review.
  
  
• You can use the EXCLUDE parameter to exclude specific objects using custom syntax. For example, EXCLUDE=INDEX:"LIKE 'DEPT%'" would exclude indexes that start with the name "DEPT".

The more I thought about the Data Pump features I liked, the longer I thought this blog would be. So here are links to three blogs I wrote on Data Pump.

• Data Pump Part 1 Features and Benefits - A general overview of the 10G Data Pump Utility.
  
  
• Data Pump Design and Architecture - We dig deeper into inner-workings of the Data Pump utility.
  
  
• Data Pump Export - We continue the Data Pump discussion by learning how to use 10G Data Pump Export to "pump" data out of an Oracle database.
  
  
• Data Pump Import - The series finale on Data Pump Import.

10G Automatic Alerting
Oracle10G provides early warning mechanisms to alert administrators of possible error conditions. The database itself collects numerous metrics that were previously collected by 9i's Oracle Enterprise Manager. Administrators are able to select from dozens of server-generated alerts. Server generated alerts are activated by a new background process called MMON. MMON is able to access the SGA directly and perform the metrics calculations and threshold monitoring.

10G Grid Control enhances these capabilities by providing administrators with a GUI interface to administer alerts and configure notification mechanisms. Administrators are able to use Grid Control or the standalone version, Database Control to set up e-mail and pager notifications. In addition, all server-generated alerts are displayed on the Database Control/Grid Control home pages.

As readers of this blog know, I'm a huge fan of 10G Grid Control. Although the database alerts can be set up in the database using the archaic command line interface, if I can do it using Grid Control with a few clicks of the mouse, that's the tool I will use. And to think I used to call 9i OEM "a Sissy GUI Tool".

Setting up 10G's "early warning system" places you in the proactive DBA category as opposed to reactive. The term we us to describe a pure reactive DBA is "firefighter". That's the DBA or DBA team that runs from one fire to the next. I will acknowledge that we are ALL firefighters from time to time. At one of my previous jobs, I spent six months working with another DBA trying to beat an absolutely atrocious third-party vendor package into submission.

You all know the type of third-party vendor package I'm talking about. Someone from your management food chain shows up in your cube and states "Hey, even though you rated this vendor an absolute 0 in all technical categories, and flatly stated that we should not consider them as a viable alternative, we chose them anyway. The Senior VP from the business unit liked the color of their screens. I'm sure you'll be able to get it working."

You then get on 72 conference calls with the vendor after the package crushes a 6 CPU box with 10 concurrent users. Most of the conversations are with the vendor's lead technical experts who are as arrogant as they are inexperienced. Their continued mantra is "you don't understand how Oracle works, it's your fault, you don't understand how Oracle works, it's your fault….". This mantra never stops even though you show them poor SQL coding, transaction looping problems, etc…

But I digress... During my career as an Oracle instructor, one of the most common recommendations I used to provide to my Oracle students was the benefits of database monitoring. My catch phrase was "We monitor by day so we don't get called at night." By the end of the DBA I class, all I would have to say is "We monitor by day…" and the class would respond accordingly. I would then award them with a set of scripts that allowed them to do database connect checks and track datafile free space.

10G Grid Control does all that (and more) for you automatically. The set of checks range the spectrum, from agent unable to connect to target (that means either the database, server or agent broke), to tablespace freespace and alert log errors. As I stated previously, 10G provides dozens of alert choices and is your one-stop-shop for proactive database monitoring. That's a good thing. I would rather wade through dozens of alerts I don't want to activate than spend ANY time coding an alert that wasn't provided by the toolset.

One of my previous blogs titled "Configuring 10G OEM Grid Control's Automatic Alert Notification System" will help you set up proactive performance monitoring in 10G. The title says it all. It provides everything you need to activate 10G Grid Control's early warning system.

Next Up
The 10 Thinks I like About 10G Continues. Could be more than 10 maybe...

We've spent a lot of time discussing SQL statement tuning. We started off this 15 blog series with a discussion on system triage and finished with Oracle access path identification and tuning. I also dedicated a dozen blogs or so to 10G Grid Control performance monitoring and analysis tools. With the abundance of long-winded information I have made available on this blog, I thought it might be beneficial to condense some of it into a review of some of my favorite tuning toolsets.

My Number One Tuning Tool of All Time - 10G Grid Control
If we are administering a 10G database, our administration tool of choice is either 10G Database Control (non-grid, standalone version), or 10G Grid Control to administer it. The advanced alerting, monitoring and administration features make this tool our preferred method of administration. The advanced performance monitoring and analysis features make it my number one tool for performance problem determination.

The future of Oracle database tuning will be administrators interpreting and implementing the recommendations generated by the intelligent advisors and ADDM. It is a foregone conclusion that Oracle will continue to improve upon its performance monitoring and analysis toolsets. Self tuning features are no longer options that are "nice to have", they are requirements for Oracle's competitive survival. SQL Server is continuing to scale, moving into areas that were once dominated by UNIX big-iron machines running Oracle databases. Oracle must compete with SQL Server's ease of use or it will undoubtedly lose market share.

As the intelligence of the advisors and ADDM increases, the need to possess an in-depth knowledge of Oracle and the usage of detailed diagnostics to improve database performance will decrease. And you heard it here first folks, I also think that reading SQL traces and statistics dumps will be a thing of the past. My crystal ball tells me that its just a matter of time until Oracle's SQL advisors make SQL traces and statistics dumps less and less important until they become totally unnecessary.

The new breed of top tuners will be the administrators who focus on how to use the toolsets and interpret their output. Not the tuners who spend the majority of time digging down into the dark, inner workings of the Oracle software. I'm not saying that knowing how the database works is immaterial. I am stating that this intimate knowledge will become less and less important as the tools mature. It's only a matter of time until tuning is relegated to a minor sub-task that can be scheduled between other activities.

Here's a listing of my favorite 10G Grid Control tools.

Host Performance Home Page
If someone tells me "the whole system is slow", the first thing I'm going to do is review the host platform's key performance indicators. 10G Grid Control's Host Performance panel provides me with exactly the information I'm looking for - current CPU, memory and disk resource utilization.

Each resource's graphical display contains links that allow me to view more detailed statistical information. Each of the drill down panels for CPU, Memory and Disk contains a drop down menu list that I can use to view current as well as historical performance statistics for the last 24 hours, 7 days and 31 days. The main performance panel also displays information on the top resource consuming processes that are currently running on the host. Once again, a drop down menu lets me switch between top CPU and top memory consumption reports.

For a more in-depth discussion on host performance tuning, please refer to my blog titled "Host Performance Monitoring Using 10G Enterprise Manager Grid Control".

10G Grid Control Database Performance Home Page
If I want to review database performance, I will access the 10G Grid Control Database Performance Home Page, which is available in both 10G Database Control and Grid Control.

The database performance home page allows me to review performance historically, identify what sessions are dominating "finite system resources", activate advisors and drill down into the top resource consumers. I can then quickly review the SQL they are executing. Does it get any better than that?

I challenge anyone to tell me that they would be able to monitor database performance more efficiently using archaic command line toolsets. By the time you have identified the problem, I'll have already contacted the user running the SQL and created a profile to improve its performance. If I have Grid Control's alerting system configured (which I always do), Grid Control's database performance threshold alert would have already notified me that a problem was occurring. By the time you got that call from an irate user, I would have had it solved. If you want to learn how to configure 10G Grid Control's performance alerting feature, please turn to my blog titled "Configuring 10G OEM Grid Control's Automatic Alert Notification System".

To learn more about 10G Grid Control's database performance monitoring and analysis features, please turn to my blog titled "Database Performance Monitoring Using 10G OEM Grid Control".

10G Grid Control Advisors
OK, I'll admit it. I'm hooked on the advisors. At first I was very wary of their recommendations, but the longer I worked with them, the better I liked them. I have lots of experience tuning Oracle, but if an advisor is going to make my job easier (and let me spend less time debugging a performance problem), I'm all for it.

For an overview of 10G Grid Control's advisory features, please refer to my blog that is aptly titled "An Overview of 10G Advisors".

Here's a listing of blogs that provide more specific information on the individual advisors. Please note that each title below also links to a blog on that advisor.

• 10G SQL Access Advisor - The SQL Access Advisor recommends a set of materialized views and indexes based on a supplied workload input. The tool can also help administrators optimize materialized views to take advantage of fast refresh and query rewrite capabilities. In addition, the SQL Access Advisor may recommend dropping indexes and materialized views that aren't being used. The SQL Advisor ranks and groups the recommendations based on the positive impact they will have if implemented.
  
  
• SQL Tuning Advisor - The SQL Tuning Advisor, as its name implies, provides SQL tuning recommendations and includes information justifying why it generated those recommendations. The recommendations may include collecting statistics on objects, new index creation, restructuring the SQL statement or the creation of a SQL profile to create a more optimal access path. I have a more detailed blog the SQL Tuning Advisor titled "SQL Access Advisor Revisited".
  
  
• Segment Advisor - Although I don't use this advisor as much as the SQL Tuning and SQL Access advisors, I do use it on a regular basis. The 10G segment advisor identifies segments that have become fragmented as a result of update and delete operations. The 10G R2 Segment Advisor has been enhanced to identify tables that suffer from excessive row chaining and row migrations. Oracle describes these objects as being sparsely populated. Not only do sparsely populated objects waste space but they can also contribute to SQL performance problems.

Automatic Database Diagnostic Monitor (ADDM)
ADDM is a "recommendation engine" that uses performance information stored in the Automatic Workload Repository (AWR) as input. Here's a blog that provides a detailed description of the Automatic Database Diagnostic Monitor. By default, AWR snapshots occur every 60 minutes. For more information on the AWR snapshots, please turn to my blog titled "Working with Automatic Workload Repository Performance Snapshots".

To learn how to tailor the 10G AWR snapshot environment to meet your specific needs, please refer to my blog titled "Configuring and Administering Automatic Workload Repository Performance Snapshots".

After the AWR information snapshot is executed, the Automatic Database Diagnostic Monitor is triggered to analyze the information contained in the AWR for the period pertaining to the last two Snapshots. ADDM's output includes a plethora of reports, charts, graphs, heartbeats and related visual aids. In general, ADDM provides users with a top-down analysis of performance bottlenecks and their associated resolutions. But you can also execute ADDM manually to generate recommendations for current, as well as historical, time periods.

Statistical information, by itself, is often useless. If you have nothing to compare a measurement to, how will you know if its good or not so good? 10G Grid Control allows two sets of snapshots to be compared to one another. The results of the comparison are displayed graphically using horizontal bars to indicate differences between the two measurements. To learn how to perform snapshot comparison, please refer to my blog titled "You Can't Improve What You Can't Compare".

Conclusion
Take some advice from your friendly ex-Oracle instructor. LEARN HOW TO USE THESE TOOLS. I can't emphasis that statement strongly enough. As I stated, ADDM and the intelligent advisors may not currently be a total replacement for DBA experience and expertise, but sooner or later, they most definitely will be. Don't get left behind - because its only a matter of time.

Learning how to tune requires that you spend time with the optimizer experimenting with different environmental settings and database objects that affect access paths. It also requires that you have a general understanding of the basics of optimization.

I dedicated an entire blog on educational resources that will help you build a firm foundation of Oracle optimization knowledge. It's critical that we understand as much as we can about optimization before we begin our experimentation. But the key to success is to spend dedicated time experimenting. I'm not professing that you spend the rest of your career becoming "at one" with the optimizer. But if you don't want to be at a total loss each time something runs long, you need to experiment to learn how to tune.

Environmental Documentation and SQL Statement Selection
Before we begin, we'll need to document some of the key parameters that affect optimization and gather some information on the data objects we will be using in our experimentation. Please refer to my blog titled "Access Path Scientific Analysis Part I" for more information on the importance of documentation as well as the documentation process itself. In a previous blog, I provided a few hints and tips on identifying what types of statements and access paths would be most beneficial to analyze.

Access Path Identification Tools
Here are the tools that I most often use to review Oracle access paths:

• SQL Scratchpad - I like to see graphical representations of access paths. Even though I have lots of experience reading SQL Trace output and the multitude of variations of access path displays, I like seeing a graphical representation. The problem is that Scratchpad must be installed on your client (comes embedded with 9I OEM) and it doesn't like accessing 10G databases. If I am tuning a 9I database and I have access to SQL Scratchpad, that is the tool I am going to use to display access path information. For more information on SQL Scratchpad and other 9I OEM tools, please turn to my blog titled "Access Path Identification Part V".
  
  
• SQL*PLUS Autotrace - If I want to quickly display the statement's access path and associated run-time performance statistics, I will use SQL*PLUS Autotrace. Here's a blog containing a few hints and tips on Autotrace.
  
  
• Explain Plan - If I don't want to run the query, and I am not concerned about the access path changing during SQL statement execution (read next bullet), I will used the tried and true "explain plan into" clause to ask Oracle to dump the access path information into a table. I'll run a statement to retrieve the data from the plan table and format the output. Please read my blogs titled "Access Path Identification Part II" and "Access Path Identification Part III" for more information.
  
  
• SQL Trace - If I want to be absolutely sure that the access path the other tools are predicting the statement will take matches what access path is chosen during runtime, I'll run SQL Trace. It is one of the only tools you can use to determine the path chosen during execution. How can the access path change during execution? Read this blog to find out! The blog contains an example of a production problem that was caused by an access path changing during runtime. For lots of information on activating SQL Trace, please turn to my blog titled "Access Path Identification Part IV". SQL Trace is the tool I will turn to if I want to analyze a set of statements that belong to a specific online transaction or batch job. The output quickly tells me what the top resource consuming statements are.
  
  
• V$SQLPLAN - V$SQLPLAN contains access path information the SQL statement takes during execution. I'll run the statement, review the access path taken during execution, make my tuning change, alter the text of the statement look different to the optimizer, retrieve the access path information again and compare the before and after results. My blog titled "Access Path Identification Part II" provides all of the information you need when you are using the V$SQLPLAN table to retrieve information on access paths taken during execution.
  
  
• Regardless of which tool I select, I will use the output to graphically display the statement's access path. Although it is time consuming, graphically displaying a statement's access path allows me to more clearly understand the access path and simplifies the tuning process. This blog will show you how to graphically display a statement's access path.

Another Graphical Display I find Useful
During my career as an Oracle instructor, I was often asked in the SQL tuning classes how I personally documented a query I was attempting to tune. This is the display that I showed them. Let's take a look at the contents, starting from the top of the page:

• SQL statement text - the full text of the SQL statement.
  
  
• Columns in the SELECT clause - a listing of the columns that are in the SELECT clause are displayed under the table they belong to.
  
  
• Columns in the WHERE clause - a listing of the columns used in the WHERE clause are displayed under the table they belong to. If the columns are tables used in a join operation, an arrow is used to designate which of the other table's columns they are joined to. If I am interested in cardinality and data skew, I would document that information directly below the column name. For more information on determining column cardinality and data skew, please turn to this blog.
  
  
• Indexed columns - The bottom of the page contains all of the indexes that are on columns in the WHERE and SELECT clauses. This allows me to easily determine if I need to create additional indexes are begin analyzing why an index isn't being used.

Performing the Scientific Analysis
Before we begin, we need to understand the effect that preloading the buffers has on SQL statement performance. If you run the query and return the results, you will be loading the buffer cache with data. As a result, subsequent runs of the same statement could be faster. This is because your statement is retrieving data from memory while the initial runs were forced to retrieve the data from disk. You'll need to run the statement twice, make your change and run the statement twice again to ensure that buffers don't impact your timings. You can then compare the output of the second execution of each test. 10G provides a SQL statement to flush the buffer cache.

If you are building indexes during your testing, don't forget to generate statistics on the new objects. You don't want the lack of object information to affect the optimization process.

In the discussion that follows, I'll be providing you with some examples of changes to make during your scientific analysis. You will run the SQL statement, record the execution time and statistics and document the access path. You will then make the suggested alteration, review the access path to see if it has changed and run the statement again to record the new performance statistics and execution time. Please note that in the examples below, I'm not going to remind you each time to run the query to evaluate the performance statistics and execution times.

Index Testing
Let's use the same sample query I used in my graphical display in this discussion:

select a.employee_id, a.last_name, b.department_id, b.department_name,
c.street_address, c.postal_code, c.city, c.state_province
from hr.employees a, hr.departments b, hr.locations c
where a.department_id=b.department_id
and b.location_id=c.location_id
and a.employee_id = 174
order by a.last_name;

There is a myriad of different changes we can make. Here's a couple of examples to get you started:

• Drop all of the indexes on the tables and review the access path. We know it will use table scans, but what join method did it choose, how about the join order?
  
  
• Build an index on one of the join columns (i.e. dept.department_id) and review the access path. Determine if it chose the index, what join method was used and the order in which the tables were joined.
  
  
• Drop that index and build an index on the other column in the join. Identify if it used the index, join method and join order.
  
  
• Build an index on employee_id, department_id and last_name. Identify if it used the index, join method and join order. Did you get an index only access path?
  
  
• Drop the index you just created and change the order putting employee_id between department_id and last_name. What was the access path?
  
  
• Change the WHERE clause to look for rows based on the last_name and first_name columns. Remove the a.employee_id = 174 clause. Build an index on the first_name and last_name columns. What happened?
  
  
• Drop the index built on the first_name and last_name columns and build two indexes, one on first_name and the other on last_name. Did the optimizer use both indexes? Which one did it choose?
  
  
• Add local predicates to the other tables and identify the access path changes.
  
  
• Identify a column in one of the tables that has low cardinality. Change the WHERE clause to access that low cardinality column. If you don't have a low cardinality column available, its pretty simple to add a column and update it with a few repeating values to create a low cardinality column. Build a bitmap index on the low cardinality column. Did the optimizer choose the bitmap index? What happens if you use a hint?
  
  
• Build a bitmap index on a column that has high cardinality (i.e. employee_id). If it doesn't use the bitmap index, use a hint to influence the optimizer to choose it. Drop the bitmap index and build a B-tree index. Does the optimizer use that index without the hint? Compare before and after execution performance statistics and timing.
  
  
• Identify columns that have skewed data. Build an index on the skewed column. Change the query to search for a value that occurs many times. Compare that execution to a search on a value that appears few times. Analyze the table and build a histogram. Re-execute the tests and compare the results.

Parameters that Affect Optimization
We also need to learn how different parameter settings affect optimization. We can start by adjusting a few of the basic parameters that have the best chance of changing the access path for a given statement. Most of the parameters can be changed in your session by using the ALTER SESSION statement. Using the ALTER SESSION statement will allow you to change the parameter for your session only. You won't have to worry about affecting other folks executing statements in the same database that you are using for your testing.

Here's a blog that provides a listing of parameters we can use during our initial testing. Please note that this is not an all inclusive list. It is a listing of parameters that I feel have the greatest chance to influence the optimizer to choose a different access path.

You need to execute the statement, change a parameter and note its affects on the statement's access path, performance statistics and runtimes. If you are going to test changing values for different parameters, don't forget to change the one you just tested back to its original setting. That way you won't have multiple changes impacting your test results. You can also test combinations of changes. Just be aware of the changes you are making.

Hints
Administrators embed hints in a SQL statement to influence the optimizer to choose a particular access path. Using hints will allow us to evaluate the affect that different access paths have on SQL statement performance. We will run the statement without any modification, review the access path and performance statistics, use a hint to (hopefully) change the access path, run the statement again and compare the before and after results. Since hints can be embedded in virtually any SQL statement, they will provide us with an easy mechanism to learn more about access paths. I have devoted a previous blog to hints. The blog also contains a demo document that will show you how to embed hints and the affects the hints have on access paths. One of the key tests is to determine the impact that operation cardinality has on SQL statement performance. You can test this by using the ORDERED hint to ask the optimizer to change the join order for statements that join more than two tables together.

Conclusion
The tests above will hopefully stimulate your creativity and allow you to generate your own test cases. As we become more experienced we'll need to review subselects, views, complex selectivity issues (and the list goes on and on…) Oracle performance tuning is truly a wonderfully complex subject. The important thing is that we start the process. That's the only way we'll become database performance scientists!

Thanks for Reading,

Chris Foot
Oracle Ace

• Let's Get Technical! - Using Deductive Reasoning and Communication Skills to Identify and Solve Performance Problems
  The first blog of this series provides some helpful hints on what questions to ask during the initial stages of application performance problem analysis.
  
  
• System Triage Part II - Host Performance Analysis Using Grid Control and Host Commands
  Using operating system commands and 10G Grid Control to monitor database server performance.
  
  
• System Triage Part III - Finding the Top Resource Consumers
  We learned how to use 10G Grid Standalone to identify the top resource consumers for a given instance.
  
  
• System Triage IV - Access Path Identification Part I
  A high-level overview of Oracle query optimization. Also discussed the difference between estimated and runtime access paths.
  
  
• System Triage V - Access Path Identification Part II
  We reviewed the two data objects that contain access path raw data - plan_table and v$sql_plan. In addition, we discussed a few of the V$ dynamic performance views that provide information pertaining to SQL statements executing in our Oracle database environment.
  
  
• Access Path Identification - Part III
  A discussion on the tools we can use to display graphical and text versions of explain plan data including the DBMS_XPLAN package, UTLXPLS.SQL, UTLXPLP.SQL and Oracle 9I Scratchpad (graphical). We also learned how to create our own graphical displays of Oracle access paths.
  
  
• Access Path Identification - Part IV
  An investigation of everyone's favorite performance analysis tools - SQL*PLUS Autotrace and SQL Trace.
  
  
• Access Path Identification - Part V
  A discussion on a couple of 9I Oracle Enterprise Manager tools that we can use to identify Oracle access paths. We also learned how we can use 9I OEM's SQL Analyze and the Virtual Index Wizard to help us better understand access paths, monitor database performance and tune statements running in an Oracle 9I database environment.
  
  
• Access Paths VI - 10G Grid Control SQL Details Panels
  We turned our attention to the 10G Grid Control's access path display panel - the SQL Details Panel. It seems like no matter where you start your investigation in 10G Grid Control, sooner or later, you'll end up with a panel that contains a link to the SQL Details Panels. We took a look at a few of the more popular ways of finding our way to them.
  
  
• Access Paths VII - Access Path Education
  Focuses on the importance of Oracle access path education. The blog contains links to resources and numerous Metalink Notes that provide access path educational information.
  
  
• Access Path Scientific Analysis Part I
  An introduction to Oracle Access Path Scientific Analysis. We learn why it is important for us to become "database performance scientists" if we want to fully understand Oracle access paths and the affect they have on SQL statement performance. We also learned how to select a test system to use as well as document the parameters and database objects that play key roles in the optimization process.
  
  
• Access Path Scientific Analysis Part II
  We learn how to use hints to influence access paths for testing purposes. Using hints allows us to evaluate the affect that different access paths have on SQL statement performance. When we finally begin our access path scientific analysis, we will run the statement without any modification, review the access path and performance statistics, use a hint to change the access path, run the statement again and compare the before and after results. Contains a document that provides a demo on using hints to influence access paths. The demo document also contains a description of some of the basic access paths that Oracle can choose from.
  
  
• Access Path Scientific Analysis Part III
  An overview of the various types of indexes as well as indexing strategies that affect Oracle access path selection. Also includes a list of parameter that can influence the optimizer to favor index access.

Selecting SQL Statements to Analyze
We need to select a few SQL statements that we can use as input to our scientific analysis. Let's take a look at the various tools we can use to select the SQL statements we can use for testing.

• Using 9I Oracle Enterprise Manager - We can use the Session Details Panel to retrieve the SQL text of statements that are currently executing in the database. The Session Details Panel also provides the access path the statement is taking. The blog on 9I OEM will show you how to drill down to find the text of the SQL statement and the access path it is taking.
  
  
• 10G Grid Control - Grid Control' s Execution Plan Panel allows administrators to display the entire text of a SQL statement and the access path it is taking. This blog on 10G Grid Control will show you how to navigate through the various panels to access the Execution Plan Panel.
  
  
• SQL Trace - If we don't have access to the "SGTs" (Sissy GUI Tools), all is not lost. We can also use the SQL Trace utility to capture the text of the SQL statement as well as its access path. To use this method, we would contact one of our friendly application developers and ask them to run a series of batch jobs or online transactions in the environment we have selected as our test bed.
  
  We can then activate SQL Trace, contact the developer to run the selected workload, deactivate the trace and format the trace's output into an easily readable format. SQL Trace provides the benefit of allowing us to capture multiple SQL statement access paths in a single execution. The blog on Autotrace and SQL Trace provides instructions on how to activate the SQL Trace utility.

What Statements Should I Choose?
We are looking for statements that utilize the different Oracle access paths. Try and find statements that use:

• Table scan and index access paths:
  
  
  • Index only - Oracle is able to read all of the data required to satisfy the query's data needs from the index structure alone.
    
    
  • Index to table - Oracle uses a row identifier to probe the table to satisfy the data request.
    
    
  • Full table scan - Oracle reads all rows from the table. Oracle will scan the table to the last block used (as opposed to the last block that actually contains data).
    
    
• Join access paths:
  Used when the statement retrieves data based on matches between two tables (i.e. retrieve all of the employees that have the department name of "Welding"). The employee information is contained in the employee table and the department information (including the department name) is in the department table.
  
  
  • Nested loop join - Good path when the join is accessing a small subset of rows.
    
    
  • Hash join - Efficient access path for joins that access larger sets of data.
    
    
  • Sort merge join - Sorts rows to allow quicker access during the join.
    
    
  • Outer joins - An outer join returns all of the rows that satisfy the particular join condition and returns additional rows from one table that do not satisfy the join condition.

We don't have to find each and every access path I outlined above. In addition, each statement that does a join will also contain table scan and/or index access paths. Try to find statements that access a single table and statements that join two, three and four tables together. It should be relatively easy to find the nested loop and hash join access paths. Sort merge and outer joins may be a little harder to find. Try and stay away from SQL statements that use parallel processing, are 13 pages long or join 5 or more tables together. We want to start with the basics. We also don't want to use statements that have exorbitantly long execution times. We'll be executing them on a fairly regular basis during testing and waiting hours for the results will lengthen the testing process.

Next Up
We devise our scientific analysis testing process.

Almost Forgot
How are the Daylight Savings Time changes treating you? I may break in one last time to discuss their impact on our organization.

Thanks for Reading,

Chris Foot
Oracle Ace

The Problem
Before we begin, we need to understand exactly what the problem is. If you haven't read my previous blogs on this topic, I highly suggest you do so before continuing.

Metalink DST Information
In my first blog on DST, I provided you with numerous links to Metalink Notes that focused on Daylight Saving Time. In my second blog, I provided a link to Oracle's new Metalink support category on DST.

Updates
We have patched dozens of databases and operating systems since I wrote that first blog on DST. I've compiled a a few notes on what we have found:

Operating System Patches

Windows
We have patched all of our customers database servers that we are responsible for providing OS support for and have experienced no issues. But some of our customers have notified us that they have experienced problems when they patched their mail and application servers.

If one implementation experiences a problem, you have a tendency to chalk that one up to that specific environment. When multiple implementations experience problems, we decided to inform our entire customer base of the potential for issues. Especially when we know that several of the customers that reported problems have top-notch Windows administrators on staff.

All of the environments had test and production operating systems that were configured to mirror each other as closely as possible. The test patches worked as advertised but a few of their production environments failed. Either the patch wasn't applied successfully or the environment refused to work after the patch was applied. In one case, Microsoft support had the administrator apply the patch again using the exact same steps and it worked the second time. Go figure. Several of the other customers were asked by Microsoft support to make various changes to their O/S configuration and attempt the patch a second time.

A few of our customers experienced mail performance problems because the DST patch caused the mail engine to generate dozens (upon dozens) of meeting time updates. The DST patch forces the mail engine to change the meeting times for all meetings that were sheduled between March 11 (new DST date) and April 1 (old DST date).

I do apologize that I don't have specifics on the the other issues. I do know that each fix was specific to that environment. I know that this may seem not very helpful at this time but I think a general warning was warranted.

Recommendations
Just because all of your test patches worked, don't assume that the production environment will also be successful. When you patch the production environment, be prepared for issues that did not occur during the test patches. You can't blame Microsoft either; test and production environments that are configured "as closely to each other as possible" aren't mirror images of each other. Close doesn't count when it comes to operating system/hardware configurations. Visit Microsoft's website regularly to ensure that there is no additional information you need to be aware of.

UNIX, LINUX
Our O/S admins experienced a few minor issues with UNIX and Linux. Please remember - if the server has its own Java engine (and you run Java programs on the O/S), your best bet is to patch both the operating system and the Java engine. We did find that several O/S vendors were updating their DST patching instructions on a regular basis.

Recommendations
Don't forget to patch that Java engine too! Visit your O/S vendor's website regularly to ensure that there is no additional information you need to be aware of.

Oracle Database Patches
The biggest issue we had was with Oracle's updates to the DST patching documentation. Since we were required to do a LOT of patching, it was important for us to get a head start. In the middle of January, we download all of the DST documentation and studied the information. We then created our plan of attack and began implementing it.

We also checked the DST area on Metalink every few days for updates. Here's where the problem arose. Some of the database patching documentation (10G specifically) can be described as "somewhat fluid" in nature. Our DST coordinator, Brian "Captain Patch" Hays found that some of the changes forced him to assemble his DST team to review their impact.

Although the updated documentation did not force us to redo any patches, we did have to revisit several of the environments that we already patched to ensure that we didn't have to reapply them. In one case, the initial documentation did not provide a complete listing of DST datatypes that could be used in table definitions. We reviewed each of the environments looking for tables containing the additional DST datatypes and were pleased to find that none were in use. I can state that when we did apply the database patches (both for the database and the Oracle Java Engine), they worked flawlessly.

Recommendations
We intend to check for updates to all Oracle DST documentation until March 11. I would highly suggest that you also do so.


Oracle Application Server
Oracle's application server was the biggest challenge. Oracle recommended numerous patches for the infrastructure database to ensure DST compliance. Many of the patches required that the infrastructure database be at a specific release. Oracle also recommended that we apply DST patches to all of the target databases that could potentially be accessed by the application server. Once again, all of the patches worked.

Recommendations
If you haven't patched your application server yet, you had better get started! Review the Metalink DST documentation for updates.

Oracle E-Business Suite Applications
All of the patches worked flawlessly.

Recommendation
Review the Metalink DST documentation for updates.

General Recommendations
Visit Oracle's and your operating system vendor's websites regularly to ensure that you have the most up-to-date information possible. Ensure that your database and application support technicians are readily available during the week of March 11th. The challenge is that some of the problems that could occur may not become readily apparent. They could very well crop up days, weeks and possibly months later. Be ready.

Thanks for reading and good luck on DST. Let me know how it works out for you!

Thanks for Reading,

Chris Foot
Oracle Ace

Introduction to Oracle Indexes
Generally, the fastest way to access Oracle data is with an index. The Oracle database contains several different indexing types that are designed to provide complementary performance functionality.

While standard B-tree indexes are most effective for columns containing a high number of different values (high selectivity), bitmapped indexes are most appropriate for columns with a relatively limited number (low selectivity) of different values. The low selectivity statement above comes with several caveats. Instead of spending this entire blog on the bitmap index/low selectivity issue, please turn to Jonathan Lewis's DBAZine article titled "Bitmap Indexes Part 1". Oracle also provides function-based indexes to allow index access using SQL that contains column manipulations in the WHERE clause.

Administrators supporting large data stores use partitioned indexes to decompose large index structures into smaller, more manageable pieces called index partitions. Starting with the 8i release, Oracle places index data in the separate index partitions based on the index's partitioning key.

Before we begin our reviewing some of the more popular Oracle index types, let me provide you with a few thoughts on indexes in general.

How Many Indexes Can I Build?
This subject has always been a matter for great debate. The DBA must balance the performance of SELECT statements with their DML (INSERT, UPDATE and DELETE) counterparts. SELECT statements that return a limited number of rows from a table yet access non-indexed columns will suffer from poor performance. Conversely, if you have too many indexes on a particular table, DML statements may be adversely affected.

The DBA must take the business requirements, application processing workload and workload scheduling into consideration when determining how many indexes to build. If you compare the performance improvements an index makes on a SELECT statement to the negative affect it has on DML statements, you will find that the benefits of building the index usually far outweigh the performance drawbacks.

Indexes on columns in the WHERE clause of SELECT statements can reduce query times by minutes and even hours. The creation of additional indexes may add additional time to on-line transactions that execute DML statements. Additional indexes will have the greatest negative impact on DML statements that access a large number of rows. The more rows that are inserted, deleted or changed, the greater the negative impact will be. Traditionally, programs that process large volumes of rows are scheduled to execute during off-hours.

The DBA must also consider the needs of the business. What process is more important to the business unit - getting the data in or getting the data out? Who complains the most? Is it the business user that must wait minutes (or hours) for their transaction or report to retrieve data or the business user that is waiting an extra few seconds for their update transaction to complete? Is the nightly batch window tight on time?

The DBA will need to find out how much time the additional indexes add to programs that process large volumes of rows. In addition, the DBA must determine when these programs run. If they run at night or do not require high-performance, consider building the index.

If the transaction update performance requirements are excessive (dot com applications are one example), keep the number of indexes to a minimum. A good recommendation is to build a set of tables that have no indexes for lighting-fast update performance and move the data to historical tables (with proper indexing) during off-hours to improve retrieval performance.

Index Monitoring
Determining if an index will increase performance is a pretty straightforward process. The administrator is focusing their tuning efforts on a particular query and is able to gather the specific information necessary to assist in the decision making process.

Dropping unused indexes is also an important part of application tuning. We learned previously that indexes force Oracle to occur additional I/O every time a row is inserted or deleted into the table they are built upon. Every update of the table's columns incurs additional I/O to all indexes defined on those columns. Unused indexes also waste space and add unnecessary administrative complexity. Since unused indexes (excluding those used to enforce integrity constraints) do not add any benefits to the environment, why keep them?

Determining if indexes were being used in releases prior to Oracle9i was a time consuming and error-prone process. EXPLAIN plan and trace output could be used but there was no single mechanism that monitored index usage at the database level.

Starting with release 9i, the Oracle database simplifies the index usage monitoring process by providing the ALTER INDEX……… MONITOR USAGE command. The statement below turns monitoring on for the index SCOTT.EMPIDX while the second statement ends the monitoring session:

ALTER INDEX scott.empidx MONITORING USAGE;
ALTER INDEX scott.empidx NOMONITORING USAGE;

The V$OBJECT_USAGE table can then be accessed to determine if the index was used during the monitoring session. When the session is started, Oracle clears the information in V$OBJECT_USAGE for the index being monitored and enters a new start time identifying when the index monitoring session started. After the index monitoring session is concluded, the USED column in the V$OBJECT_USAGE table will contain the value 'YES' if the index was used during the monitoring session and the value 'NO' if it was not.

Parameters that Impact Index Usage
The parameters listed below influence the Oracle cost-based optimizer to favor or not favor index access. Please note that this is not an all inclusive list. It is a listing of parameters that I feel have the greatest chance to influence the optimizer to choose, or not choose, index access paths. We'll use these parameters during our scientific analysis.

• OPTIMIZER_MODE = first_rows or first_rows_nnnn - The optimizer chooses the best plan for fast delivery of the first few rows or the first nnnn rows. The first_rows_nnn replaces the first_rows parameter in later Oracle releases. The first_rows is available for backward compatibility. More often than not, that access path will include an index. This optimizer mode tends to favor nested loop joins over hash and merge scan.
  
  It is important to note that using this mode is not a switch. It won't change each and every table scan and hash join to index access and the nested loop join method. The optimizer, at times, will favor index access and nested loop joins. The reverse goes for the all_rows optimization mode below.
  
  
• OPTIMIZER_MODE = all_rows - The optimizer chooses the best plan for fast delivery of all of the rows that queries return. The optimizer may decide to choose a full table scan over index access and hash joins instead of nested loop.
  
  
• OPTIMIZER_INDEX_COST_ADJ = xxxx - This parameter lets you tune the optimizer to be more or less index "friendly." It allows the administrators to influence the optimizer to make it more or less prone to selecting an index access path over a full table scan.
  
  The default for this parameter is 100. This setting tells the optimizer to evaluate index access paths at the regular cost. Any other value makes the optimizer evaluate the access path at that percentage of the regular cost. For example, a setting of 50 makes the index access path look half as expensive as normal.
  
  
• OPTIMIZER_INDEX_CACHING = xxxx - You set this parameter to a value between 0 and 100 to indicate the percentage of the index blocks the optimizer should assume are in the cache. Setting this parameter to a higher value makes the indexes on the inner table of a nested loop joins look less expensive to the optimizer. The end result is that the optimizer may for favor a nested loop join using an index.
  

Index Affects on Access Paths
If a table only contains a few hundred rows, queries may run faster if the optimizer chooses to read all of the blocks in the table as opposed to using an index. The I/O generated traversing the index blocks to get to the table row entries would be higher than if Oracle read just the blocks allocated to the table being accessed.

What if we access a larger table using a column with poor selectivity? Selectivity describes the number of different values stored in a column. Poor selectivity means that the column has many values that are the same. If our statement contains a WHERE clause that searches for a column value that is contained in 90% of the table's rows, it is best that Oracle, once again, read each and every row in that table. Conversely if a WHERE clause searches for a column value that appears in 1% of the table rows, it would be beneficial for the optimizer to choose an index.

Here are some examples of index access paths that we will see during our testing:

• Index only - Oracle is able to read all of the data required to satisfy the query's data needs from the index structure alone.
  
  
• Index to table - Oracle searches through the index structure looking for the key value(s), then uses row identifiers to probe the table to satisfy the data request.
  
  
• Index unique scans - The SQL statement accesses an index using a column (or columns) that are defined in a unique or primary key index with an equality condition.
  
  
• Index range scans - Oracle scans a set of entries in the index to satisfy a query.
  
  
• Index skip scans - Oracle is able to break down a multi-column index and view them as smaller subindexes. This is achieved by Oracle "skipping" the leading columns in the index and using columns that appear later in the index's definition.
  
  
• Full scans - Oracle scans all of the index entries. Kind of like a tablescan on an index. Oracle drops down to the leaf blocks and traverses the leaf blocks using the leaf pointers.
  
  
• Fast full scan - Oracle uses multi-block reads to read both leaf and non-leaf blocks. Non-leaf (branch blocks) are discarded.

But indexes influence more than just index-related access paths. Indexes can also impact the type of join operations used. Here's an "over the top" example to clarify. If you are joining two tables together in a SQL statement on join columns that have good selectivity (returns relatively few rows compared to the table size), Oracle will favor index access paths and a nested loop join. If you don't have indexes on the join columns, Oracle may choose table scans using a hash join instead of the nested loop join method.

Index Types
Let's continue our discussion by reviewing some of the more popular types of indexes: B-Tree, Bitmap and Function.

B-Tree Indexes
A traditional B-Tree index stores the key values and pointers in an inverted tree structure. The key to good B-Tree index performance is to build the index on columns having a lot of different values. Oracle describes this as "good selectivity" Oracle is able to quickly bypass rows that do not meet the search criteria when searching through indexes built on columns having a high degree of selectivity.

Bitmap Indexes
As I stated previously in this blog, there is a sense of confusion about bitmap indexes and the benefits they provide to columns with low selectivity. Instead of spending this entire blog on the bitmap index/low selectivity issue, please turn to Jonathan Lewis's DBAZine article titled "Bitmap Indexes Part 1".

The optimizer can be stubborn at times. It can be particularly stubborn when you want it to choose a single bitmapped index for an access path. A single bitmap index may not be chosen at all. The optimizer will be more inclined to choose bitmapped indexes as an access path if it can use multiple bitmapped indexes simultaneously. That's where the benefits of bitmaps are realized.

Bitmap Indexes and Concurrency
Anyone accustomed to database programming understands the potential for concurrency problems. When one application program tries to update data that is in the process of being changed by another, the DBMS must sometimes forbid access until the modification is complete in order to ensure data integrity.

Each entry in a B-Tree index entry contains a single ROWID. When the index entry is locked during an update, a single row is affected. A bitmap lock affects a range of entries which could have a negative impact on other transactions attempting to update rows already locked.

Locking issues affect data manipulation operations in Oracle. As a result, bitmapped indexes are not appropriate for OLTP applications that have a high level of concurrent insert, update and delete operations. Concurrency is usually not an issue in a data warehousing environment where the data is maintained by bulk loads, inserts and updates.

Function Based Indexes
Oracle 8i solved an indexing problem that had been affecting database performance for close to a decade. Before Oracle8i, any SQL statement that contained a function or expression on the columns being searched on in the WHERE clause could not use an index.

For example, the statement:

SELECT * FROM employee_table
WHERE Upper(first_name) = 'CHRIS';

would not use an index. A full table scan would be required to retrieve the desired result set. We now know that we are able to use B-tree and bitmap indexes to speed query performance.

In Oracle8i and later releases, we are able to build both bitmap and B-tree indexes on columns containing the aforementioned functions or expressions. The following index could be used to increase performance of the query:

CREATE INDEX upper_first_name on employee_table (upper(first_name));

Function based indexes will affect the performance of DML statements that manipulate the columns contained in the function based index. The more complex of the expression used, the more time the database will require to update the index.

Wrapup
The intent of this blog was not to provide you with an all-inclusive education on Oracle indexes. There are folks that spend dozens of hours learning about access paths and the affects that indexes, hints, statistics and parameters have on Oracle optimization. We are laying the groundwork to begin our own scientific analysis of Oracle access paths. As I stated previously, we need to begin experimenting on our own to fully understand the Oracle optimization process. In my next blog, we'll use the information we learned in this series to start our experimentation.

SQL Hints
Administrators embed hints in a SQL statement to influence the optimizer to choose a particular access path.

By using hints, you are telling Oracle that your access path is better than the one the optimizer is choosing. It’s a safe assumption that most of us aren’t as smart as the optimizer. Let it make the choice, unless you are certain the optimizer is choosing the incorrect access path.

But what happens if the optimizer is making incorrect decisions? Before you begin adding hints to SQL or freezing access paths using Optimizer Plan Stability or 10G Profiles, consider taking the following steps first:

• Determine if it is actually an incorrect access path that is causing the performance problem. It may be some external influence affecting the SQL (hardware, workload, and so on).
  
  
• Identify and review the SQL taking the bad access path for proper SQL coding techniques.
  
  
• Verify that statistics have been generated on the tables and indexed columns. The Oracle-supplied utility DBMS_STATS is currently the preferred method of collecting statistics.
  
  
• Review parameters that affect SQL optimization (optimizer_mode, optimizer_index_cost_adj, optimizer_index_caching, optimizer_dynamic_sampling, optimizer_features_enable, optimizer_max_permutations).
  
  
• Investigate system statistics. Is it activated? Is it configured correctly if it is activated? Should it be activated?
  
  
• Does the application use bind variables? If so, investigate bind peeking quirks.
  
  
• Check for skewed data. Consider using histograms to compensate.
  
  
• Go to Metalink and review optimization bugs for your release. Oracle could have already identified your issue and fixed it.
  
  OK, so you have performed all of the actions cited previously and you find that the optimizer is actually making an incorrect decision. Regardless of what some industry pundits may tell you, the optimizer is NOT infallible; it can make mistakes. Oracle created hints for a reason, and wouldn’t have made them public if it didn’t think we really needed them from time to time. If you are forced to add hints to the query to improve its performance, do so intelligently and judiciously.

Using Hints to Compare Oracle Access Paths
OK, now that I have provided you with my standard warning on hints, the intent of this blog is to learn how to use hints to influence access paths for testing purposes. Using hints will allow us to evaluate the affect that different access paths have on SQL statement performance. We will run the statement without any modification, review the access path and performance statistics, use a hint to (hopefully) change the access path, run the statement again and compare the before and after results. Since hints can be embedded in virtually any SQL statement, they will provide us with an easy mechanism to learn more about access paths. We are on our way to becoming database performance scientists!

We'll begin our access path scientific analysis by using a very basic set of hints to influence the optimizer to choose a different access path. The hints I will be using in my introductory demo are:

• Hints for optimization mode - We will be asking Oracle to optimize the statement using different optimization goals. Since we are using Oracle9i for my demo, we'll be asking it to use choose, first_rows, all_rows and rule.
  
  
• Hints for access paths - Access path hints ask the optimizer to choose the access path it recommends. We'll be asking Oracle to use an index that it didn't choose in the original access path it generated. We'll also be asking the optimizer to choose a full table scan instead of using an index.
  
  
• Hints for join operations - Oracle provides several different join methods for statements that join one, or more, tables together. We'll ask the optimizer to choose nested loop, merge scan and hash joins.
  
  
• Hints for join order - Oracle only joins two tables at a time. If multiple tables are joined, join order also describes the overall order of the tables being accessed. Oracle will join two tables and create an intermediate result set which is then used as input to the next join.
  
  Join order plays a significant role in query performance. Both in the outer and inner tables selected and the overall join order. In general, you want to reduce the number of rows processed as soon as you can during the processing of a given SQL statement. The sooner you can reduce the number of rows being sent to future operations, the faster the query will usually run. We'll ask the optimizer to choose different join orders to determine the impact it has on SQL performance.

The bullets above are just a subset of all of the hints that are available. For a complete listing (and their definitions), please refer to the Oracle Database Performance and Tuning Guide for your release. You'll find the documentation on Oracle's Technet website.

Recommended Toolsets
The two tools that I most often use to review SQL performance during my own scientific analysis is SQL*PLUS Autotrace and SQL Trace. Please refer to my blog titled "Access Path Identification - Part IV" to learn more about using SQL*PLUS Autotrace and SQL Trace utilities to evaluate SQL access paths and SQL statement performance.

Demo Document
The intent of this demo is not to train you to identify which access path is most optimal for a given situation. Its intent is to help you gain experience interacting with the Oracle optimizer. Take it from your friendly ex-Oracle instructor, spending time experimenting with the optimizer and analyzing the performance statistics that different access paths generate is critical to your tuning education. There really is no substitute for time spent "in the seat" performing your own scientific analysis on query optimization.

The demo will show you how to use hints and the ALTER SESSION SQL statement to influence the optimizer to take a different access path than it normally would. You can then compare the the access paths and their associated performance statistics to obtain a better understanding of what accces path is best for your test queries.

Here's a link to the demo document. It is in Word format for ease of reading.

In my next blog, we'll discuss what SQL statements to use for your testing and what to look for when you compare the results.

Thanks for reading.

Chris Foot
Oracle Ace

I have found that most good tuners share a common set of traits. They are inquisitive by nature, understand that there is no substitute for experience and dedicate lots of time performing scientific analysis on SQL statement and database performance.

One of my favorite tuning gurus is Jonathan Lewis. Jonathan used terms like "scientific analysis" in his DBAZine podcast. It's really a very appropriate description of his activities. After reading many of his works, I would describe him as a database performance scientist. He identifies something he wants to learn more about, creates a test environment, executes a series of tests, analyzes the data and provides documented results to reienforce his theories.

That's what we all MUST DO to expand our knowledge on the Oracle optimization process. We need to become database performance scientists. We do that by creating a test environment, running and documenting performance baselines, changing the environment to influence the statement's access path and documenting the results.

We'll begin this series by learning how to select a test environment and document it. In upcoming blogs, we'll discuss the tools we will use to measure and compare our results, select a set of SQL statements to use for our testing, create our test cases, execute the tests and document the results.

Setting up a Test Environment
Running test cases to identify the affects that different database environmental settings have on a statement's access path is not as monumental as it may seem.

Most production databases have a test counterpart. Select an environment that is actively used by your application developers. Don't worry, if one of our test causes a statement to "run longer than anticipated", we can use the trusty ALTER command to kill our session. We can also run our workloads during lull times. Lastly, the majority of changes we will make to influence the statement's access path will be done in our own session. We won't be making changes that affect the entire database environment.

I prefer to use an active test environment because it allows me to easily select SQL statements to use as test cases, the data is usually more representative of what will be found in production and the developers will most likely have a firm understanding of the active workload being executed.

You need to talk to the developers who are responsible for running workloads on that test environment to ensure that the changes being made to the test data doesn't skew your results from one test execution to the next. You can't run a test, have a load insert another 100 thousand rows in the table, run another test and expect to have a good comparison. You want your test bed to be active but not so active that it complicates your testing process or causes your tests to generate incorrect results. You'll need to use common sense when selecting the test environment as well as determining the most optimal time to run your test cases.

We also want to choose a test environment that is providing adequate performance for the workloads being executed upon it. We really don't want to use an environment that isn't performing well to begin with.

There's a myriad of options available to you. If you don't want to impact any of your test environments, create a test environment of your own. Clone one of your test databases to a sand-box environment if you can.

You'll also want to make sure that statistics are up to date on the database you will be using as a test bed. If you are running 10G databases, the database will run statistics jobs for you automatically (isn't 10G great?). If you don't have statistics run automatically, it will be up to you to analyze the data objects to ensure that the statistics optimally represent what is stored in the data structures.

Documenting the Parameters That Affect the Optimization Process
The next step is to document the environment. There are a couple of dozen parameters that affect optimization and SQL statement performance. To begin, we are going to choose the basic parameters that are easy to change and have the biggest impact on optimization. These are not all of the parameters that can influence the optimization process, just the ones that are easy to change and provide the best chance of successfully achieving an access path change.

It is important that we read the documentation beforehand for these parameters for the specific Oracle release that we are using as our test environment. We know that each Oracle release may contain enhancements to these parameters that change the affect they have on the optimization process and how we alter them to different values.

We'll want to document the following parameters to begin our scientific analysis:

• cursor_sharing - For our first set of initial tests, we'll hardcode values in our selection criteria to ensure that our statements aren't affected by cursor sharing. To learn more about cursor sharing and the impact it has on statements using bind variables, please turn to my blog titled "System Triage IV - Access Path Identification Part I"
  
  
• db_file_multiblock_read_count - The number of blocks that Oracle will read in one I/O when performing a sequential scan of data (i.e. table scan). We also need to note that if we are using Oracle 10.2 as our test environment, the database itself may adjust this parameter dynamically if it identifies that the buffer cache is being flooded by blocks being retrieved using a table scan.
  
  
• optimizer_features_enable - This parameter allows you to make the optimizer behave as it would for a specific release of the Oracle database. You set the value to a release identifier (the listing of the optimizer releases that you can set this parameter to is provided in the Reference Manual) and the optimizer will act as it would if the database were at that release.
  
  
• optimizer_index_caching - Administrators will often set this parameter in conjunction with optimizer_index_cost_adj to influence the optimizer to use more indexes and Nested Loop joins. Setting this parameter makes Nested Loop joins look less expensive when compared to Hash or Sort-Merge joins.
  
  
• optimizer_index_cost_adj - Allows the administrator to make index access more, or less, costly to the optimizer. The default value of 100 tells the optimizer to evaluate the index using the regular cost. The lower you set this value, the less costly the indexes will look to the optimizer. Conversely, the higher you sent this parameter, the more expensive the indexes will look.
  
  
• optimizer_mode - Sets the approach the optimizer will use when analyzing queries. Since there have been a few changes made between Oracle 9i and Oracle10g, I'll provide information on both sets.
  
  

  Oracle 9i provides the following settings:
  

• Rule - The optimizer will use the rule based approach during analysis. The rule based optimizer does not use data statistics as input when generating the access path. Each access path is assigned a numerical ranking. The rule based optimizer chooses the access path that has the most favorable numerical ranking. The rule based optimizer has been superceded by the cost based approach. There are a few cases where I have seen the rule based optimizer choose a better access path than the cost based method - but not many.
  
  
• Choose - If optimizer_mode is set to choose, the optimizer is able to switch between rule and cost-based optimizations. When optimizer_mode is set to CHOOSE, the optimizer uses the all_rows cost-based approach for a SQL statement if there are statistics in the dictionary for at least one table accessed in the statement. If you generate statistics on one table, every query that accesses that table will use the cost-based optimizer. What happens if other tables in the query do not have statistics collected? The optimizer will make an educated guess on the statistics for those tables. The problem is that Oracle isn't always a good statistics guesser and the end-result is a "less than optimal" access path. To learn more about the affect that statistics have on cost based optimization, please refer to my blog titled "System Triage IV - Access Path Identification Part I".
  
  
• first_rows - Influences the optimizer to choose an access path that minimizes response time. Most often used for online transaction processing systems that return small result sets. The optimizer favors Nested Loop joins and index access.
  
  
• all_rows - Influences the optimizer to choose an access path that minimizes total execution time. Most often used for decision support and data warehouse environments. The optimizer tends to favor full table scans, Hash and Merge-Scan joins.
  
  

  Oracle10g settings for optimizer_mode:
  

• first_rows - The same influence on the optimizer as it did in Oracle9i.
  
  
• first_rows_n - Where "n" = 1, 10, 100, 1000. Influences the optimizer to optimize queries to provide the fastest response when returning the "n" number of rows. Acts as a throttle, which allows you to better balance the optimization process.
  
  
• all_rows - The same influence on the optimizer as it did in Oracle9i.

We will use the ALTER SESSION SQL statement to alter these parameters during our scientific analysis on the affects they have on the Oracle optimization process.

Documenting Our Test Tables
After we document some of the parameters that affect optimization, let's turn our attention to documenting the tables we will be accessing. In my next blog, I'll provide you with a few hints and tips on how to select or create SQL statements to use in your test cases but let's continue our discussion on documentation.

The following information will provide you with a good base of information on the data objects our statments will be accessing. Since we are just beginning our scientific analysis, we'll use basic storage objects (tables and b-tree indexes). We'll discuss some of the more complex objects (bitmap indexes, partitioning, etc.) in later blogs.

• Row counts for all tables that are accessed by our test queries. Can be found in the num_rows column in dba_tables if we have statistics generated for our tables.
  
  
• Number of blocks the table is using. Can be found in the blocks column in dba_tables if we have statistics generated for our tables.
  
  
• Index listing for all indexes on our tables. The query below will provide you with a listing of indexes for a given table:
  

  select b.index_name, b.column_name, a.uniqueness, b.column_position
  from sys.dba_indexes a, sys.dba_ind_columns b
  where a.index_name = b.index_name
  and a.table_owner='&table_owner' and
  a.table_name = '&table_name'
  order by b.index_name, b.column_position
  /

• Once we find all of the indexes and columns in those indexes, let's check them for both selectivity and data skew. Selectivity is the number of unique values for a given column. Skew means that some values can occur a few times in a column while other values can occur many, many times. Since data skew will affect optimization, that information will also be important to us.
  
  We can easily find the selectivity for a single or multi-column index by accessing the distinct_keys column in our dba_indexes table if we have statistics generated. For multi-column indexes, we will want to check the individual selectivity for each column in our multi-column index. We can do this with the following query:

  select count (distinct index_colname) from owner.table_name;
  
  Where index_colname is one of the columns in our multi-column index,owner is the table owner and table_name is the name of our table. We'll need to do this for all columns in our multi-column indexes.

  We'll use the following queries to identify data skew and determine if any histograms are on our tables.

  select index_colname, count(*) from owner.table_name
  group by index_colname;

  Where index_colname is the column name, owner is the table owner and table_name is the name of our table. We'll need to do this for all columns in our indexes. We'll do this for both single column and multi-column indexes.

  select * from dba_tab_histograms where owner='&owner' and table_name='&table_name';
  

  Where owner is the table owner and table_name is the name of our tables. We'll need to do this for all tables that our queries will be accessing.

Coming Up Next
We'll continue our discussion in the next blog when we review the hints we will be using to influence the optimizer's choice of access paths. In addition, we'll discuss the various tools we will use to measure and compare our results.

Thanks for Reading,

Chris Foot
Oracle Ace

Important New Information
In that last blog, I provided you with numerous links to Metalink Notes that focused on Daylight Saving Time. Since then, super-DBA Mark Shore informed me that Oracle has created a new DST support category in Metalink's Knowledge Base.

It's pretty easy to find the new DST support area. After you log on to Metalink, you'll notice that Oracle provides a list of navigation tabs at the top of each page.

Click on the second navigation tab from the left titled "Knowledge". Metalink will respond by displaying the Knowledge Browser home page. Scroll half-way down the page until you find the section titled "New in the Knowledge Base."

At the end of that paragraph, there will be a link titled "USA DST 2007 Category". If you click on that link, Metalink will respond by displaying the Daylight Savings Time (DST) USA 2007 Support Category home page.

Daylight Savings Time (DST) USA 2007 Support Category
The new support category in Metalink's Knowledge Base contains links to ALL of the notes on Daylight Saving Time. It is your one stop shop for DST information. If you scroll down to the Topic Listing Table, you'll see that the first note (Note:412789.1) provides information on Oracle DST Webcasts.

Oracle DST Webcasts
We have already attended the DST Webcasts and have found them to be very informative. Oracle provides DST Webcasts for all of their products that are affected by the 2007 DST changes:

• Webcast 1 - Oracle Database and Client. Provides information on the Oracle database, Oracle client software, Oracle Enterprise Manager and the Oracle JVM.
  
  
• Webcast 2 - Oracle E-Business Suite Applications. Provides information on all of the components affected by DST including the data server and the middle-tier services, such as Forms, Concurrent Manager, Apache, JServ and the JVM. The presenter will tell you which components are affected, which aren't, and what to do to prevent the 2007 DST changes from becoming a problem.
  
  
• Webcast 3 - Peoplesoft Application. Describes the changes to Peoplesoft that will be required to support the 2007 DST changes.

The remaining DST Webcasts combine the Application and Database DST Webcasts into a single presentation. All of the Webcasts begin with a presentation on DST and finish with a Q&A session. We found the Q&A sessions to be just as helpful as the presentations.

If you haven't attended one of the DST Webcasts, all is not lost. There are a few dates and times still available. In addition, Note:412789.1 contains information on how to replay prior DST Webcasts. Oracle also provides you with a couple of different options to choose from. You can replay a prior DST Webcast by accessing it straight from your browser or you can download it to your PC and replay it at your leisure. I highly suggest that you replay the DST Webcasts or attend one of the future sessions.

Our Status
I thought it may be beneficial to provide you with a status on our own internal 2007 DST project:

• Since we support both the database and E-Business Suite applications, we have created two teams that focus on 2007 DST changes. Each team member is required to become an expert on the impact that the changes will have on the components they are responsible for. This includes identifying what needs to be patched, learning the patching process and coordinating patching efforts with their fellow technicians and our customers.
  
  
• All of our technicians have read the Metalink Notes pertaining to DST and have attended one, or more, of the DST Webcasts.
  
  
• We executed the scripts provided in Metalink Note:412971.1 and have determined that none of the databases we are responsible for supporting use the TIMESTAMP WITH TIMEZONE datatypes.
  
  
• We executed the scripts provided in Metalink Note: 397770.1 to identify all of the databases that have the Oracle JVM installed. We will be applying the JVM patch to all Oracle 8.1.7, 9x and 10x databases that support applications using Java programs.
  
  
• We will be applying both the timezone and JVM patch to all 10G databases.
  
  
• We are patching all Oracle E-Business Suite applications.
  
  
• All Grid Control databases will have both the JVM and timezone patch applied.
  
  
• We have sent notifications to all customers to review all third-party applications, middle-tier products, operating system and external Java engines to ensure that they are 2007 DST compliant.

Your Status?
Twenty some years ago, Craig Mullins and I had the good fortune of being trained and mentored by someone who I would describe as being one of the top managers in the database profession.

We once had to revert to a previous backup in DB2 because an operator had allowed one of the current backup tapes to be overwritten. Since DB2 is intelligent enough to go back to a prior version without problem, the impact on the recovery was minimal. He called me to his office and asked "Who is ultimately responsible for the availability, reliability and recoverability of the databases here?". I stated that I was. He then calmly declared "Then you should understand that since you are ultimately responsible, you are also responsible for ensuring that there are no weak links in the chain."

I continue to live by that rule. Its blatantly obvious that we need to ensure that our databases, administration toolsets and applications are 2007 DST compliant. But we also need to play a dominant role in our organizations to ensure that all of the components (operating systems, third-party applications, middle-tier products, external Java engines, etc.) that interact with our databases are also 2007 DST ready. I am confident that my organization has done everything we can to ensure the transition is smooth.

My challenge to you is "Is your shop ready for 2007 DST?.

Thanks for Reading,

Chris Foot
Oracle Ace

To improve energy conservation, the Energy Policy Act of 2005, Pub. L. no. 109-58, 119 Stat 594 (2005) changed the start and end dates of Daylight Time. Starting in 2007, Daylight Time begins on the second Sunday in March and ends on the first Sunday in November. That means that in 2007, Daylight Time begins on March 11 and ends on November 4. In 2008, Daylight Time begins on March 9 and ends on November 2.

Impact on the Oracle Database Ecosystem
Most computer systems have been configured to accept the Daylight Time changes that have been in effect since 1986. My personal desktop computer pops up a box in April and October to let me know that it needs to change the date to adhere to the Daylight Time policy.

If I don't apply the recommended Microsoft fixes, my PC will have the incorrect time for the duration of the new DST 2007 change periods (March 11, 2007 - April 1, 2007 and between October 28, 2007 and November 04, 2007). I'm assuming that if I manually change the time and don't apply the fix, I'll wake up one morning to find that my PC operating system made another change for me using the old Daylight Saving Time dates. One way or another, my PC's date will be off.

You understand how this could have a pretty dramatic effect on virtually every database ecosystem that we support.

Warnings and Recommendations
I am by no means a 2007 DST impact expert. Personally, I find a lot of the documentation provided by the vendors on this subject to be vague and confusing. I'm providing you with links to assist you in your own education.

We are taking the safe route. When in doubt, we are asking the vendor detailed questions to clarify their statements. Our feeling is that this is what we pay support for - to get answers to questions. Many of the vendors responses will tell you to "play it safe" and patch the systems.

Operating Systems and Java Engines
If your organization does not apply the operating system patches to correct the timezone change, the timestamps those systems use will probably be incorrect by one hour for the duration of the new DST 2007 change periods (March 11, 2007 - April 1, 2007 and between October 28, 2007 and November 04, 2007). That means any dates and times that the database pulls from the operating system for SYSDATE and DATE will be incorrect. In addition, the operating system's Java engines are also affected because they use their own embedded time zone information.

Recommendation - Visit your operating system provider's website to determine what patches are required for the operating system and the Java engine.

Oracle Database Impact
Oracle is providing two patches, one for the JVM and one for the time zone files. Each patch has a different set of requirements on when to implement it.

Oracle JVM
Like its operating system counterpart, Oracle's embedded Java engine uses embedded time zone information and is impacted by the new Daylight Time start and stop dates.

Here's a snippet from one of Oracle's white papers:

JVM patch - for all Oracle versions 8.1.7, 9.x, 10.x (database only):
If you have the Oracle JVM installed you are advised to apply the JVM patch. See section "JVM Fixes" below for more details.

The problem is that a lot of shops install the Oracle JVM when they install the database but don't run Java programs in Oracle. I'll provide you with a Metalink Note that will provide you with instructions on how to determine if the Oracle JVM is installed and what to do if it is.

Time Zone Files
Depending on the release, there are Oracle datatypes that will be affected by the Daylight Time changes. Oracle states that the TIMESTAMP and DATE datatypes do not have time zone knowledge. But the 10G TIMESTAMP WITH LOCAL TIME ZONE (TSLTZ) and TIMESTAMP WITH TIME ZONE (TSTZ) datatypes and the TZ_OFFSET function take their time zone information from Oracle's time zone files. These datatypes can be used as column types or as PL/SQL types.

The time zone files are installed during the database installation. Oracle states that the version 3 and higher time zone files include the 2007 Daylight Time changes. I'll provide you with a Metalink Note that will show you how to determine what version you have.

Here's the killer, Oracle is stating that if you use the affected datatypes, you will be required to patch the database and all clients that need to access those datatypes. Once again, read the notes I am providing you and make your own decision on what to do.

Oracle Grid Control and Agents
Oracle Grid Control agents must have their times synchronized with the operating system. If not, the agents will not start. Both the O/S and agents must be patched or not-patched in order for Grid Control Agents to work. Oracle has supplied a manual work-around in case of emergency. The repository databases must also be patched. We have asked Oracle to provide us with clarification on the 9i OEM agents. We are patching all 9i OEM and 10G Grid Control repositories.

Oracle E-Business Suite Applications and Application Server
I have included an Oracle Metalink noted that discuss the impact the Daylight Saving Time changes have on Oracle applications and the Oracle Application Server.

Recommendation for all Oracle products - Visit the Metalink website, use the Metalink notes I provide as a starting point.

Third-Party Applications
There are numerous applications that will require a patch in addition to the required operating system patches.

Recommendation- Your organization will need to contact all third-party applications to determine the impact that timezone changes have on your applications.

Oracle Metalink Notes
The following Metalink Notes will provide you with a good starting point to begin your 2007 Daylight Time changes education. Good luck!

Start your education by reading these notes:
Note 359145.1 - Impact and overview of 2007 USA daylight saving changes on the Oracle database.

Note 402742.1 - USA 2007 DST Changes: Frequently Asked Questions for Oracle RDBMS. Probably the most beneficial of all the overview documents.

Note 397281.1 - USA 2007 Daylight Saving Time (DST) Compliance for Oracle Server Technologies Products.

Generic time zone information:
Note 357056.1 - Impact of changes to daylight saving time (DST) rules on the Oracle database.

Note 340512.1 - Timestamps & time zones - Frequently Asked Questions.

Note 396906.1 - Patches to Update the Time Zone Transition Rules.

Note 412160.1 - Updated Time Zones in Oracle Time Zone File patches.

Installing and using the utltzuv2.sql script:
Note 406410.1 - Dealing with existing Database time zone use when updating Oracle TZ-files.

Note 396670.1 - Usage of utltzuv2.sql before updating time zone files in Oracle 9.

Note 396671.1 - Usage of utltzuv2.sql before updating time zone files in Oracle 10.

Note 397013.1 - Advanced use of utltzuv2.sql.

Note 402614.1 - Time Zone Data in the Data Dictionary During a Time Zone File Update.

Note 399832.1 - Workarounds when utltzuv2.sql patches are not available for your patchset.

Applying the time zone file patch:
Note 359145.1 - Impact of 2007 USA daylight saving changes on the Oracle database.

Note 396387.1 - Workarounds when Database time zone patches are not available for your patchset.

Note 396426.1 - Effects on client and middle-tier of applying time zone patches on the Oracle Database.

Note 402316.1 - Database Time Zone Patching Strategy in the Enterprise.

Platform specific information for those platforms with some special patching issues:
Note 406196.1 - VMS: DST 2007: Information for OpenVMS platforms.

JVM issues:
Note 359145.1 - Impact of 2007 USA daylight saving changes on the Oracle database.

Note 397770.1 - How to check if Oracle JVM is installed in the Database.

Note:276554.1 - How to Reload the JVM in 10.1.0.X and 10.2.0.X.

Note:209870.1 - How to Reload the JVM in 9.2.0.X.

Oracle Applications:
Note 403659.1 2007 - Daylight Saving Time Changes For Oracle Applications.

Note:403311.1 -Impact of U.S. 2007 Timezone Changes on Oracle E-Business Suite, Release 11i Environments.

Grid Control Issues:
Note:401834.1 - EM Agent will not start - due to a Daylight Savings change - (Workaround). Explains why it is impacted, how to fix it and the workaround.

Thanks for Reading,

Chris Foot
Oracle Ace

Gaining a well-rounded and deep understanding of Oracle access paths and SQL performance is a wonderfully challenging task. One of the hurdles is the time required to learn how to tune SQL. DBAs are being challenged with ever-increasing workloads, shrinking DBA staffs and increasingly complex technologies to support.

In addition, the majority of shops view database administrators as much more than just "table jockeys." The DBA is often seen as the go-to technician because of their traditionally strong problem solving skills. The DBA is also viewed as the IT staff's technical generalist because of the working knowledge they have in many different facets of information technology. Those of us that have been working in this profession for any time at all understand that the term "database administration" is really a misnomer. We have to know everything from application and data design to network communications and operating systems (and everything in-between).

As a result, many of us don't have enough time to dig deep into the many different facets of the Oracle database environment. Oracle has recognized this issue and has created the various advisors in 10G to assist us in the monitoring and tuning process. I have written dozens of blogs on the toolsets provided in 10G that are designed to reduce the amount of time we spend administering, troubleshooting and tuning the Oracle Ecosystem (database, operating system, hardware server).

The Importance of Understanding Oracle Access Paths
That being said, all DBAs must have a firm understanding of Oracle access paths and SQL tuning. During my career, I've learned that there is nothing that can drag down an Oracle Ecosystem than poorly performing SQL statements.

That's when management suddenly doesn't care about how much you know about RAC, standby databases and data design. They want the database "fixed" and running smoothly again. They suddenly have a single-minded purpose. They quickly begin the chant, the chant that can only be described as the management mantra - "is it fixed yet, is it fixed yet, is it fixed yet, is it fixed yet, is it fixed yet….."

I've actually had my entire chain of command stand behind me in order of where they fit into the management food chain (team lead, project manager, section manager, division manager, VP…) while I was working on a database performance issue. I turned around and chuckled when I saw the lineup of managers. Funny, they didn't share my humor.

You need to get the performance problem fixed, you're nervous and you have a 15 page SQL statement on your screen. This is where a strong education in Oracle access paths becomes "somewhat handy."

SQL Tuning Topics You Must Understand
Here's a quick laundry list of topics you'll need to know. All of them are important. I have tried to include resources later in this blog that will provide you with information on all of them. You can use this listing to check off the topics during the education process.

• The Oracle release's impact on optimization and SQL tuning. Each new release contains features that affect access paths and SQL performance. Sometimes good and sometimes not so good.
  
  
• Oracle parameters that affect the optimizer. There are a handful or two of startup parameters that can influence the SQL optimization process. I have provided several articles below on this topic.
  
  
• Optimizer modes - Rule, choose, first_rows, all_rows. Each of the modes influence the optimizer to create access paths for the type of workload the database is responsible for supporting. For example, the first_rows optimizer mode may be OK for online transaction processing (read a record/write a record) but probably won't generate efficient access paths for a data warehouse database where millions of records are summarized.
  
  
• Oracle data statistics - How the optimizer uses them, how they affect access paths.
  
  
• Oracle system statistics - Later releases of Oracle can also incorporate the system load information during optimization.
  
  
• Basic access paths
  • Index only - Oracle is able to read all of the data required to satisfy the query's data needs from the index structure alone.
  • Index to table - Oracle uses a row identifier to probe the table to satisfy the data request. Why read all of the rows in a table if you can use an index structure to retrieve just the rows you need?
  • Full table scan - Oracle reads all rows from the table. If the statement is going to read the majority of a table's rows, why would you want it to needlessly traverse an index to get the data? You are reading extra index blocks for no reason. You also need to learn the impact that the high-water mark has on full table scans. Oracle will scan the table to the last block used (as opposed to the last block that actually contains data).
    
    
• Join access paths - Used when the statement retrieves data based on matches between two tables (i.e. retrieve all of the employees that have the department name of "Welding"). The employee information is contained in the employee table and the department information (including the department name) is in the department table.
  • Nested loop join - Good path when the join is accessing a small subset of rows.
  • Hash join - Efficient access path for joins that access larger sets of data.
  • Sort merge join - Sorts rows to allow quicker access during the join.
  • Cartesian join - The tables being joined do not have join clauses that relate the two tables together.
  • Outer joins - An outer join returns all of the rows that satisfy the particular join condition and returns additional rows from one table that do not satisfy the join condition.
    
    
• Join order - Oracle only joins two tables at a time. If multiple tables are joined, join order also describes the overall order of the tables being accessed. Oracle will join two tables and create an intermediate result set which is then used as input to the next join.
  
  Join order plays a significant role in query performance. In general, you want to reduce the number of rows processed as soon as you can during the processing of a given SQL statement. The sooner you can reduce the number of rows being sent to future operations, the faster the query will usually run.
  
  
• Subqueries - A select within a select statement. Can be one of the trickier statements to tune, especially when you have multiple subqueries embedded within each other.
  
  
• Indexes and selectivity
  • B-tree indexes are good for column(s) that have many unique values (high selectivity)
  • Bitmap indexes are used for column(s) that do not have many unique values (low selectivity)
  • How SQL statement predicates can determine if an index can be used. There are times when the way a statement is coded prevents Oracle from choosing an index as the access path. A common problem that often leads to poor performance. You do get a chance to flog the application developer responsible, though.
    
    
• Types of index access paths
  • Index unique scans - The SQL statement accesses an index using a column (or columns) that are defined in a unique or primary key index with an equality condition.
  • Index range scans - Oracle scans a set of entries in the index to satisfy a query.
  • Index skip scans - Oracle is able to break down a multi-column index and view them as smaller subindexes. This is achieved by Oracle "skipping" the leading columns in the index and using columns that appear later in the index's definition.
  • Full scans - Oracle scans all of the index entries. Kind of like a tablescan on an index.
  • Fast full scan - Oracle uses multi-block reads to retrieve the index blocks.
    
    
• Sorting - Many operations require the database to sort a result set. It could be that the query wants to return the data in a particular order (ORDER BY, GROUP BY). In addition, some joins require that the data be sorted during the operation's execution. You'll need to understand why sorts are performed and the impact they have on performance.
  
  
• Views - Views can really complicate the tuning process. You think you are accessing a few tables in a query and then find that you are actually joining views together that also contain join operations.
  
  
• Hints - My next blog will show you how to use hints to influence access paths and how hints can be used to educate yourself on the performance of a particular access path operation.
  
  
• Bind variables and bind peeking - In the first and second blogs of this series, I described the impact that bind peeking can have on SQL statement optimization. Bind peeking may lead to the predicted access path not matching the access path taken during execution.
  
  
• Query transformation - Oracle can rewrite a query during the optimization process. Learn how Oracle uses view merging, predicate pushing, OR expansion and subquery unnesting to attempt to improve execution performance. This will occur on a regular basis and you need to understand how query transformation works and how it affects SQL performance.
  
  
• Local predicates vs join predicates - A local predicate accesses a bind or hardcoded variable (i.e. emp_id = :empid, emp_id = 13344) while a join predicate is used to join two tables together (i.e. emp.dept_id = dept.dept_id)
  
  
• Predicate usage - You need to understand how predicate usage affects indexes and access path generation. I have provided information below that discusses predicate conditions (=, >, <, etc) as well as how predicates affect index utilization and access paths.
  
  
• Operation cardinality - The number of rows returned by a particular access path operation. Importance of operation cardinality is magnified as the number of tables accessed in the query increases. As stated previously, the sooner you can reduce the rows sent to future operations, the better your query will perform.
  
  
• Skewed data and histograms - What happens when you have an index built upon a column in a million row table that has twenty occurrences of the value "OUT OF STOCK" and the rest of the column values contain the value "IN STOCK". This is an extreme case, but the impact is that even though you may access the table looking for an "OUT OF STOCK" value, Oracle will most likely perform a table scan. You'll be searching close to a million values, while you need to retrieve only twenty of them. An index would be a much better access path, but Oracle sees that the column has such poor selectivity that it won't choose it. A histogram identifies skewed data and is able to provide the optimizer with the information it needs to make a more educated decision when choosing between a table scan and index access.
  
  
• Parallel processing and partitioning - A best practice for large data stores is to partition data into smaller subsets of data. Partitioning allows data to be broken down into these smaller subsets yet still be viewed as a single-entity by the application. Parallel processing breaks a single request for data into one or more processes that access the data in parallel and return the data to the calling application.
  
  
• Parsing - Learn the differences between hard parses vs soft parses. When the application sends a statement to the database for processing, one of the first steps in the execution process is called a parse. Oracle will check the statement's syntax, check security, generate the access path, etc.. Like most operations the less steps it needs to perform the better. You will need to understand the impact that bind variables have on the parsing process and how the hard parse/soft parse ratio affects query and database performance.

Access Path Education
You can start your education on the different access paths that are available to the optimizer by reading Oracle's Database Performance Tuning Guide that is provided in each Oracle release's documentation. Before you buy third-party books on any topic, I highly suggest that you read Oracle's documentation first. The importance of this suggestion bears repeating - READ ORACLE's DOCUMENTATION FIRST. Here's a link to the 10G Database Performance Tuning Guide.

You'll need to create an account, but its free to register and the process is painless. Virtually every topic that I described above is covered in the Database Performance Tuning Guide. It is very important that you read the guide that pertains to the release that you are working with. Each release contains enhancements to the optimizer as well as new features that affect the optimization process.

Oracle Classroom Education
OK, since I'm an ex-Oracle Instructor, you could have predicted that my next recommendation was to sign-up for an Oracle class on SQL performance tuning. Here's a previous blog that provides you with a few hints and tips to obtain the most from your classroom experience.

Oracle Education offers a class that focuses on SQL tuning. Here's an excerpt from the class description provided on the Oracle Education Website:

What you will learn:
--------------------------------------------------------------------------------
This course is designed to give the experienced SQL Developer or DBA a firm foundation in SQL tuning techniques. The participant learns the necessary knowledge and skills to effectively tune SQL in the Oracle Database 10g. They learn about tuning methodology as well proactive tuning and reactive tuning methods. Students are introduced to the benefits of the new automatic tuning mechanisms available in Oracle Database 10g. On completion of the course they are able to compare and contrast the steps involved to tune manually as in prior releases as well as use the automatic SQL tuning features provided in the current release. Students gain a thorough conceptual understanding of the Oracle Optimizer, and reinforce instructor-led learning with structured hands-on practices. The course uses a series of challenge-level workshops, allowing students to "play, discover, and learn" at their own level and pace. The students learn to use the Oracle diagnostic tools and facilities: Automatic SQL Tuning components, EXPLAIN, SQL Trace and TKPROF, SQL*Plus AUTOTRACE. Students also learn to influence the behavior of the Optimizer by changing the physical schema and modifying SQL statement syntax.

Metalink Notes
Oracle's premier web support service is available to all customers who have current support service contracts. Oracle MetaLink allows customers to log and track service requests. Metalink also allows users to search Oracle's support and bug databases. The website contains a patch and patchset download area, product availability, product life-cycle information and technical libraries containing whitepapers and informational documents. A few of the white papers and notes that pertain to SQL tuning, optimization and access paths are provided below (do a search using the note number on the main page in Metalink to retrieve the note):

• 199083.1 - Query Tuning Overview - Lots of good links to other articles.
• 398838.1 Frequently Asked Questions. Very good discussion on the optimizer. Discusses queries not using indexes.
• 248971.1 - Query tuning best practices. LOTS of links to other notes. Links discuss a wide range of topics. From parameters that affect optimization to system statistics impact on optimization.
• 35934.1 - Common Issues and Misconceptions about the cost based optimizer.
• 68735.1 Diagnostics for Query tuning.
• 67522.1 - Diagnosing why a query doesn't use an index. Very helpful.
• 207434.1 - Tuning Queries - Quick and Dirty Solution.
• 372431.1 - Troubleshooting a new queries.
• 154354.1 - Using the FIRST_ROWS for fast query response.
• 100229.1 - Measuring Index Selectivity.
• 41954.1 - Hash join operations - a little dated but still valid.

Third Party Books
If you want to learn how the optimizer works, I highly suggest that you read Jonathan Lewis's book titled Cost-Based Oracle Fundamentals. It is one of the most educational and informative books I have ever read on the cost-based optimizer. Here's my review of the book for DBAZine. If I could only recommend one book on the Oracle optimizer, Jonathan's book would be it. Jonathan also maintains a blog that focuses on a wide range of topics but does include a lot of discussions on Oracle tuning.

Tom Kyte's Ask Tom Website also provides a lot of information on SQL tuning, access paths and proper coding techniques. One of Tom's trademarks is to use a snippet of code to reinforce the information he is conveying. I'm a big fan of using examples and a big fan of Tom's website.

Articles and Presentations
Article by Don Burleson on cost based optimization. Discusses parameters that affect the cost based optimizer. Statistics, helpful hints and tips.

Advanced CBO Article and Presentation. Advanced discussion on how statistics influence the optimizer, why the optimizer doesn't always take the correct access path. Written By Wolfgang Breitling.

Article on the Oracle 10G Costing model changes and how they influence access paths. Listing of parameters that affect optimization. Author is Kimberly Floss.

Oracle Statistics Article - System statistics affects on the optimizer. Very good article that provides lots of details. Written by Jonathan Lewis

Article on why Oracle won't use an index (when you think it should). Excellent article written by Jonathan Lewis.

Test, Test, Test
Experience pays. You need to spend time "in the seat" learning how to tune. Read the above information and find a database that you can use as a test environment. You need to work with tables that have small numbers of rows and queries that return small result sets. You also need to run queries that access tables with high numbers of rows and return large result sets.

I'll show you how to use hints to change access paths in then next blog. Change the access path with a hint and run the query. Influence the optimizer to take a different access path and join order and keep a tuning record of the changes and runtimes.

Don't' fall into the trap of favoring one access path or join method for all situations. I once overheard a conversation between a developer that just moved from an online transaction system to the data warehouse team and the data warehouse DBA. The developer was looking at an access path and stated "I hate hash joins." The warehouse DBA stated "Not in this environment you won't". All access paths and join methods have a place in Oracle optimization. Its up to you to learn which ones apply for a given situation.

Thanks for Reading,

Chris Foot
Oracle Ace

All Roads in 10G Grid Control Lead to the SQL Details Panels
10G Grid Control's primary display tool for providing information on specific SQL statements is the SQL Details Panels. It seems like no matter where you start your investigation in 10G Grid Control, sooner or later, you'll end up with a panel that contains a link to the SQL Details Panels. Let's take a look at a few of the more popular ways of finding our way to them.

Database Performance Home Page
The Database Performance Home page provides a couple of links that lead to the SQL Details Panels. If we start from the top, we'll see the Sessions: Waiting and Working Chart in the middle of the screen. This display allows us to determine exactly what resource our disgruntled users are waiting for.

Clicking on one of the colored areas of the Sessions: Waiting and Working Chart allows us to drill down to the specific resource being waited on. Clicking on any of the colored areas will display a drill down panel that provides details on that specific resource. For example, if we click on the purple I/O color on our chart, we'll navigate to the I/O drill down page.

No matter what resource you drill down into, they all have the same look and feel. All of the resource panels display the particular resource's historical utilization and two pie charts on the bottom of the screen that allow administrators to drill down into the top sessions and SQL statements utilizing that resource. If we click on the links listed under the Top Waiting SQL: User I/O heading, 10G Grid Control will respond by displaying the SQL Details Panels. If we click on the links under the Top Waiting Sessions: User I/O heading, 10G Grid Control will display the Session Information panel. The Session Information panel contains a link to the current SQL being executed which is displayed using the SQL Details Panels.

If you are interested in learning more about the Database Performance Home Page, please turn to my blog titled "Database Tuning Using Oracle 10G Grid Control".

Top Sessions Panel
If we return to our Database Performance Home Page, we need to scroll down to see the Top Sessions and Top SQL links. When we click on the Top Sessions link, 10G Grid Control will respond by displaying the Top Session display panel. If you look at the red box, you'll see that the Top Sessions panel is actually part of the Top Consumers set of display panels.

We navigate to the Session Details Panels for each session in our report by clicking on the link under the SID column heading. The Session Details Panel provides a link to the current SQL being executed which 10G Grid control displays using the SQL Details Panels.

Top SQL Panel
Let's return to our Database Performance Home Page. The link below Top Sessions is the Top SQL. Clicking on the link displays the Top SQL panel. The Top SQL panel has to be one of my favorite display panels in 10G Grid Control because it allows us to view SQL performance historically. The Top SQL panel provides a listing of SQL Details links that provide us with access to the SQL Details Panel.

Viewing Performance Historically
10G Grid Control allows us to take the database "back in time" and review performance historically. This screenshot shows the Database Performance Home Page in historical mode. Historical mode is activated by the drop down menu (red box) that is displayed in the upper right hand corner of the page.

The historical mode display provides a slider that allows us to go back in time and view performance statistics captured during a specific time period. When I drag the slider to a time in the past, 10G Grid control displays the Database Performance Home page statistics for the chosen time period. If we click on the Top SQL link at the bottom of the page, 10G Grid Control responds by displaying the Top SQL Period display panel. This panel provides links to the SQL Details panels for the top SQL statements that were active during the chosen time period.

SQL Details Panels
Now that we have learned how to navigate to the SQL Details Panels, let's take a look at the information they provide. The first screen displayed is the Execution Plan panel.

Execution Plan
The Execution Plan panel provides us with the SQL statement's access path taken during execution. The top section will display the entire text of the SQL statement. The next section displays where the panel retrieved this information from (cursor cache), the optimizer mode, data capture time, plan hash value and optimizer mode.

The bottom section provides us with all of the basic information we need to identify the access path this query is taking. It also includes key performance indicators including the number of rows that were used in each step, the time to perform the step and CPU and I/O cost indicators.

Current Statistics
Let's take a look at the Current Statistics Panel and the wealth of tuning information it provides. The panel will tell us if this SQL statement is being influenced by a SQL Profile. We'll have an entire discussion on SQL Profiles in an upcoming blog. But for the sake of this discussion, a SQL Profile is a stored access path that we can generate using the SQL Tuning Advisor.

Time Model information provides key performance indicators including elapsed time per execution, CPU time per execution and the statement's wait ratio. The Execution Statistics includes the number of executions, parse calls, rows per fetch, rows per execution and executions per parse call.

Execution History
The Execution History panel provides us with another key set of performance indicators. It shows us the statement's performance historically. The top chart displays the number of seconds the statement took to execute. This allows us to quickly determine if a statement's spike in elapsed and CPU times can be attributed to an event that occurred during a specific time in the past.

The Executions chart shows us how many times a statement was executed historically. This allows us to determine if an overall degradation in database performance can be attributed to an increase in the number of executions of this statement. I have used this panel to find a problem in a program that was causing it to loop and execute a particular statement millions of times.

Tuning History
The Tuning History Panel displays information on past tuning exercises for this statement. If you have generated any SQL Profiles, this panel will display information on each SQL Profile generated. It will also provide information on which SQL Profile is currently being used. The panel allows users to activate and deactivate each of the SQL Profiles that were generated for this statement. As I stated, we'll focus on SQL Profiles in an upcoming blog and I'll show you how to use this panel to switch between the different SQL Profiles created for a given statement.

Wrapup
I hope you enjoyed this blog on the SQL Details panels. I think you'll agree that it provides us with a good set of information to begin our SQL performance tuning activities.

Thanks for Reading,

Chris Foot
Oracle Ace