I am getting about a 99% skip ratio for f5 data models that do not complete. The searches take quite some time to summarize the datamodels and I need to adjust the settings most likely. What I am concerned about is that this is a prebuilt datamodel for the f5 application and I do not want to modify any of the settings that will conflict with the application.
What can I do to fix this without messing up the application?
The maximum number of concurrent auto-summarization searches on this instance has been reached 624828 97.33 %
The maximum number of concurrent historical scheduled searches on this instance has been reached 17113 2.67 %
I increased the number of searches per core to 4 and I don't have any performance problems with that regard. What settings do I need to change now to help fix this issue?
The F5 app accelerates 50 models by default. Make sure you need all of those.
Basically every 5 minutes it is trying to run 3 acceleration searches times 50 models over three month period. That is absurd. I would start by disabling all of them and looking for what data you actually have and/or care about and selectively re-enable them as needed.
Also consider going into advanced acceleration settings and reducing the concurrent acceleration searches to 1 if you need to enable a bunch of these.