Hi team, @DerekLaw
We have a 6‑machine ArcGIS Server cluster that is under heavy load, but it is also well resourced. We are not observing any CPU or memory pressure, RAM usage is consistently at or just below 80%, and CPU utilisation remains within acceptable limits.
One of our feature services, hosted in an Enterprise Geodatabase (SQL Server), is heavily used and is currently handling approximately 140 requests per second max. From an ArcGIS Server perspective, the platform is handling this load without issue.
However, ArcGIS Monitor is consistently failing for this particular ArcGIS Server cluster. We operate multiple ArcGIS Server clusters, and this issue appears isolated to the busiest one. When reviewing both Windows Event Logs and ArcGIS Server logs, we are not seeing any meaningful errors or indicators that explain the Monitor failures.
This raises the question:
Is ArcGIS Monitor known to struggle or fail when monitoring very busy ArcGIS Server clusters, even when system resources are not constrained? if so, is there a way to remediate this?


