High Number of Criticl Alerts in LIVE
Our application is in LIVE and of late we are observing a large number of Critical Alerts in the logs on a Day to Day basis. Any average day of operation throws up 500 Critical Errors on an average with PEGA0026,PEGA0042,PEGA0004 being the pre dominant ones (more than 98%).
Most of these errors are pointing to the connection pool settings which might be too low when compared to the actual number of connections being requested. We are looking at altering these settings and arriving at a reasonable value so that such errors are reduced to Zero.
Have any of you faced similar issues? Any suggestion/s comments on possible remediations?
**Moderation Team has archived post**
This post has been archived for educational purposes. Contents and links will no longer be updated. If you have the same/similar question, please write a new post.
Updated: 29 Jan 2016 1:09 EST
Cognizant
IN
Assuming you have a websphere + Oracle setup ...
It takes around 30 ms to establish a DB connection based on my observation.
If you have PEGA0026 alerts for 300ms ( and above) it is a clear indication of poor DB connection pool configuration.
Some tips below for stopping this alert ...
- DB Connection pooling should always be done at application server with JNDI reference to access the connections; never use prconfig.xml to connect to DB.
- Leverage Type 4 thin drives as much as possible.
- Set max connections to 30, keeping this too low / high does not help.
- Leverage "Idle timeout" setting if you have not already done so.
- Disable unnecessary agents in your application , PDN has an article on this.
Cognizant
IN
https://community.pega.com/video-library/pega-developers-conference-2013-pega-7-architecture-depth (1H:12M:07S)
