Advertisement

06.30.2003 at 08:59AM PDT, ID: 20664353
[x]
Attachment Details

Oracle XA Performance Tuning

[x]
The Solution Rating System

With so many solutions, how can you tell which solutions are most likely to help you and which ones are not? To provide you with a tool to use, we rate our solutions based on various elements that most accurately determine if a solution is a quality solution. To explain what factors affect the solution rating, here are the elements we take into consideration when formulating our solution rating.

  • The Grade of the Solution
  • The Zone Rank of the Expert Providing the Solution
  • The Number of Author and Expert Comments
  • The Number of Experts Contributing
  • The Feedback of the Community

Your Input Matters
Because of the way the system is set up, the most important variable in this equation is you. As a member of Experts Exchange, you are able to cast your vote on the quality of the solutions in regard to how complete, accurate, helpful and easy to understand each solution is. When you provide your feedback, each rating is adjusted accordingly. So, if you see a solution that has a poor rating that you think is a good solution, let us know by rating it. As you do, the rating will be adjusted and will become more accurate for other members of our site.

If you have any suggestions that you would like to make for our rating system, please ask a question in the Suggestions Zone of Community Support.

Thank you!
7.8
Zone:

Oracle Database

Tags:

oracle, xa
Does anyone have any performance tuning recommendations for Oracle XA using the Oracle JDBC driver ?  I compared the performance of running a J2EE application using Oracle non-XA thin driver and Oracle thin driver with XA.

The difference is very significant - XA is roughly 3 to 5 times worse than non-XA, and the difference as far as utilization is an increase %CPU of Oracle.  Looking at the difference from Oracle using statspack, the run with XA spent a lot more time on "buffer busy waits" on data blocks + additional processing time spent on:
JAVA_XA.xa_prepare_new (:2,:3,:4);
JAVA_XA.xa_commit_new (:2,:3,:4,:5);
JAVA_XA.xa_start_new(:2,:3,:4,:5,:6);
JAVA_XA.xa_end_new(:2,:3,:4,:5);
 
Each of these  PL/SQL procedures are taking up 15-18% of total CPU time of Oracle.  They are actually java stored procedures when I looked ath the JAVA_XA package and they seem to be very expensive in tems of CPU.

I tried running the Oracle OCI driver to use native XA so that I dont spend CPU time in these JAVA_XA package but I ran into  XAER_RMERR runtime exceptions when I specify nativeXA=true.

I got nativeXA to work with WebLogic's Oracle JDBC jDriver and I was able to bypass the JAVA_XA packages and lower database CPU utilization but I was still having very poor performance.  Oracle statspack showed that I spent 95% of the time waiting for "buffer busy waits" on data blocks.  I was not getting these waits when I run my application without ORacle XA.  

So given this, my questions are:

1) Can anyone recommend some tuning information for Oracle XA ?

2) Why do I get "buffer busy waits" with XA but not without XA ?  

thanks,
Tyrone



 
Start your 7 day free trial to see this answer
Answered By: Vinay_dba
Expert Since: 06/12/2003
Accepted Solutions: 94
Vinay_dba has been an Expert for 5 years 7 months, during which he has posted 306 comments and answered 94 questions. Vinay_dba is just one of 1067 experts in the Oracle Database Zone. 1 expert collaborated on this answer, which was graded an "A" by the asker.
 
 
20081119-EE-VQP-47