Opennms UI response is very slow after applying constant load
Description
The opennms UI is very slow after constant load is applied through the minions. i.e. the page response time is very slow. Steps to recreate: 1. apply a load of 3000 syslog messages per second for 20 minutes 2. open opennnms events list and refresh the page. 3. Navigate to opennms resource graphs and notice the page response time of events is very slow and the graphs do not load
This was fixed by managing queue sizes and is further addressed by the sink pattern changes. Marking as fixed.
Seth Leger November 30, 2016 at 4:10 PM
OK, sounds good. I'll mark that issue as a dependency.
Tim Fite November 30, 2016 at 2:39 PM
Seth - this issue was logged before we added queue sizes on the seda queues. I believe this was caused by the seda queue filling up the heap. I know we verified there were OOM Excpetions and I believe this has since been fixed with the seda queue size setting. I think once the PR is done and you verify it can be closed.
Seth Leger November 30, 2016 at 2:35 PM
Can you collect more information about the system's performance and attach it to this bug? It would be useful to have:
The opennms UI is very slow after constant load is applied through the minions.
i.e. the page response time is very slow.
Steps to recreate:
1. apply a load of 3000 syslog messages per second for 20 minutes
2. open opennnms events list and refresh the page.
3. Navigate to opennms resource graphs and
notice the page response time of events is very slow
and the graphs do not load