Good afternoon. I’ve been monitoring my server for a while now, and have been keeping an eye on crash events and non-crash events. So far I’ve not found many correlations to any particular events, or CPU spikes. Sometimes the CPU spikes are there, sometimes they are not. There remain few clues as to what “triggers” the Out-Of-Memory server crashes, but instead it seems that, as the server’s age progresses, it something just randomly pushes it over the edge. So – still inconclusive there.
I’m still wondering if there might be any explanation for why the non-heap memory and metaspace might be on an ever-increasing trajectory, and why, as @carehart mentioned, this would not be considered a problem.
There is no unexpected increase in traffic or any unusual requests. It’s just a random moment, and the server says, “we’re done.”
So the only thing I’m left to ponder is the ever-increasing metaspace and non-heap space, which continue their ceaseless march upward throughout the life of the server.
Why might metaspace and non-heap space always increase? Is this considered normal? Or symptoms of an underlying issue?