800xA Basic History -OPC Log Data
Also observed that in basic history Provider metrics "LogMgrQueueLenth" value is High(Refer attached screen shot).
After restarting, the basic history value changed to zero, then trend are updating very fast.
Why the value is increasing ? Any recommendations and configuration?
If you see long logmgr queue with PercentSynchronized at 100, the friction is coming from outside, i.e. trend clients.
Use the Find Tool, search for Aspect Category=Log Template, then Log Configuration and drag the results to the Consistency Checker Tool and have them all screened and if necessary, corrected.
The number of logs in the image is very small considered the limit of 10.000 logs per service group of Basic History.
Even if the IM's stagger function hasn't been applied (resulting in IM requests arriving in large, or even very large (all logs) chunks) the queue should not spike so much given the size of the config. Make sure the IM's Sample Blocking Rate is a magnitude or more larger when compared to the source log's Sample Interval. Do not use a SBR ratio of 1:1. Stagger can fix this for you more or less with no effort from you. Make sure you learn this tool and use it to flatten the workload IM passes onto Basic History.
What clients does this historian serve? Lots of operator trends in process graphics, mini (embedded) PG2 trends, etc? Heavy reports digging extensively into the logged data scheduled too often?
With Basic History, no disk is too fast, but many are too slow. SSD or PCIe is better than 7200 RPM SATA.