Hallett_German

CA Tuesday Tip: 25 APM CE Tips for Stable and Scalable Sites

Discussion created by Hallett_German Employee on Feb 9, 2013
Latest reply on Feb 14, 2013 by MaryGreening
CA Wily Tuesday Tip by Hallett German, Sr. Support Engineer for 2/12/2013

According to my notes, this is the 25th Tuesday Tip that I have written. A lot of underdocumented areas and best practices have been covered so far and many more are planned. Suggestions for future topics are welcomed.

Here are twenty-five tips that you can do to have a stable and scalable APM CE environment. Many of these were covered in past Tuesday Tips.

1. Do your own self-heathcheck of APM CE environment on at least a bi-monthly basis. Things to check are the following:
- APM CE metrics being/closed to being clamped.
- TIM issues due to changes in traffic load and quality.
- Private keys about to expire.
- User groups near limi.t
- TIM/TESS-MOM GUI are less responsiveness.
- Too many defects/incidents/users.
- New errors in logs.
- New types of APM CE logs.
- Review perflog and supportability metrics for EM and database health.

This can be supplemented by a bi-monthly/quarterly health check by APM Professional Service staff.

2. Maintain a configuration book with screenshots of APM CE settings. This can be supplemented by a runbook.

3. Maintain a spreadsheet of the values matched on for transaction definitions. This will help avoid definition overlap.

4. Create and maintain a monitoring strategy document.

5. Create and maintain an APM monitoring roadmap which lists new/updated applications will be monitored that period.

6. Have multiple APM environments including one to test real production scenarios and to capture updated APM definitions.

7. In production, have a business service for sandbox testing before deploying an APM transaction definition.

8. Have a naming standard for APM CE elements. See the tech note on this site for additional information.

9. Do an architecture review at least twice a year for fast changing environments.

10. Use the Run to CSV tool periodically to see if definitions are no longer seeing hits.

11. Have a separate APM environment for pre-APM 9/historical data.

12. Be able to recreate environment using backup, configuration and business transaction export.

13. Create a disposal environment when it make sense such as for temporary reports and applications. (See past Tuesday tip.)

14. Keep the environment clean of unneeded and used information. (See past Tuesday tip.)

15. Document your APM architecture and network environment and keep it current. This will be a tremendous help during problem resolution.

16. For transaction definitions, typically have no more than twenty transactions per application. When setting up transaction matching rules for no more than one to four parameters. Remember that this is an AND not OR condition so each parameter-value pair must be matched.

17. Continuously optimize and maintain your transaction definitions. Else old definitions will result in false positives or missing data may occur.

18. Determine your APM CE Administration Philosophy and then choose appropriate tasks to perform. (See past Tuesday Tip.)

19. Make APM as part of the development lifecycle so new/updated APM CE definitions are added to production before deployment.

20. Always keep your servers time synched, with plenty of disk space, CPU, and Java heap.

21. Do not capture anything that is not needed such as users/user groups, applications, report, and metrics. Also do not keep anything longer than needed.

22. If an outage does occur, do an after-action review to determine which changes are needed and how this can be proactively tracked in the future.

23. When time allows, read the various documents available on this site to take advantage of current and new features.

24. Periodically discuss with APM users on what they like and like less about the product to determine any missed opportunities or future projects that can be pursued.

25. Internally issue a quarterly APM Report Card on what changes were made in the environment, what problems were resolved due to APM, APM operating issues encountered, and discussion of APM roadmap/monitoring strategy.


I hope to meet or talk to you in the future and would love to hear back on how Tuesday Tips have helped you!

These are the discussion questions for this article:
1. How of these twenty-five tips are you doing?
2. Do you have additional tips to recommend?
3. How has the Tuesday Tips articles helped you?

Attachments

Outcomes