Backtrace & Buffer overrun

[htech]

I already posted this on Reddit but no one responded so maybe I will have more luck here... I need help understanding what is happening with my ZM system. I have 8 cameras that are 2688x1520 and the same 8 cameras connected to their second stream of 640x360, for a total of 16 monitors connected to a ZM server running as a VM in ESXI. The server has 14GB of ram, and 10vCPUs which works out to about 21GHz. After about five minutes I start to get errors about buffers overrunning, then zoneminder crashes and needs to be restarted. the ram usage is around 80%, and the average load for the 10 cpus is 9~11. the errors that I usually get are either related to backtrace errors or buffer overrun. I'm not sure what error logs would be useful so please let me know.

[rockedge]

have you tried to increase the level of debug logging in ZM? Options->Logging->LOG_DEBUG selected to ON then increase the
Options->Logging->LOG_DEBUG_LEVEL to something like 5 or 6 to start off. then check /var/log/zm

[htech]

So now I have got debug info but i'm not really sure what to do with it.

[rockedge]

what do you have?

[htech]

I have a bunch of debug info, what is important?

[bbunge]

What is the space used in /dev/shm? With those cameras at that resolution the tmpfs could be filling up and maybe causing issues with your database server. Speaking of which have you checked the space in innodb buffer pool?

[alabamatoy]

Speaking of which have you checked the space in innodb buffer pool?

How would one go about checking this?