Anyone else have issues with AppInsight for Exchange chewing through a ton of PowerShell sessions? Default for WinRM is 5 per user. I upped it to 10. SAM ate through that. I upped to 20. SAM ate through that. Annoyed, I upped it to 50 and it ATE THROUGH THAT! You'll see a whole pile of wsmprovhost processes running for the SW service account. Sometimes it hovers in the 10-15 process range but sometimes it flies up to the 40-50 range and stays there. Each process eats about 100-200MB of RAM so I've seen the whole mess chewing upwards of 8GB of RAM. I can restart WinRM to flush it all out but they just come back. I have a case open [627842] but was curious if anyone else has run into this and possibly found a fix?
Also, as a bonus, a few days ago one of my nodes running the template completely stopped reporting in what appears to be most, but not all, of the PowerShell pieces of the template. The ones that don't work show up as "unknown" in SAM. The service monitors appear to not work either, all come back grey.
Oddly tho, things like Replication Status Checks works fine and that *is* a PowerShell monitor.
I also get this oddity:
Bit of a mixed duck here. The other MBX server running it also has the same "SW uses a whole freakin ton of shells" issue but the monitoring still works for the most part. It occasionally will smack the 50 limit and fail a poll but picks up immediately again at next poll [or if I go dump all the wsmprovhost processes it has spawned on the server].