Forum Discussion
Joe_Williams
2 years agoProfessor
Our setup will be different due to being a MSP, but we have collectors with upwards of 140k instances. It is all about adjusting and tuning the collector.
One of the things we do is run an automation that deletes interfaces that have had a status 2 alert for more than 30 days. Most of these are interfaces that aren’t around any longer.
We have built a standard around collector sizing and threadpool counts. That has helped a lot. We have also noticed that if a collector is maxing itself out, more memory actually hurts for some reason. More cpu helps a crap ton.
Related Content
- 4 months ago