Hi,
Since a few weeks we have a memory issue, it rises up to 97%, we need to disable “historian instance” every 7-8 days. A disk writing problem?
n.b Server memory 256 GB, PMU connected 150
Regards
Pietro
Is it possible that the disk is full?
Note that by default the system uses the C:\ as a temp drive and will need enough space here to move files into long term storage location.
Also, this can be configured to be another drive besides C:\
Temp storage and Long Term storage are in the same disk, free disk 1 TB.
Regards
Pietro
Is the historian server working okay besides the memory buildup? Do the timestamps of any of the real-time streams gradually fall behind as the memory builds up? Can you query the latest data from the archive itself?
Many of the reported memory leaks are caused by a module in the system that is, for whatever reason, consuming data faster than it can process. If you go into the installation directory for openHistorian, there is a file called Status.txt. That file is updated by the openHistorian every 30 minutes and has detailed information about the status of adapters defined in the system. If you search for the word “queued”, you will find metrics about the size of various queues used throughout the system that might be building up data. If any of those numbers are high, that may point you in the right direction in determining why memory might be building up. If none of them are high, then it’s probably something else entirely.