While we troubleshoot CSV Latency counter in case # 02464332 with Kirill we found that VeeamOne collects performance data incorrectly.
Now VeeamOne collects performance data for CSV File system object only from one node. If look deeper VeeamOne connects to cluster name (e.g. cluster.company.com) that owned by one of nodes in that cluster (e.g. node1.company.com) and collects performance data only from this node. That what Kirill told me, and that is incorrect.
Corresponding to Microsoft article (https://blogs.msdn.microsoft.com/cluste ... -counters/) performance counters must be collected and aggregated across all cluster nodes.
Kirill also told me that this actually not a bug, that is design of VeeamOne. And to correct this is needed to make a post on this forum, so it may be will fixed in future releases.The most important consideration when looking at counters on a CSV is to keep in mind that values of the counters is not aggregated across nodes. For instance if one node tells you that “Avg. Disk Queue Length” is 10, and another tells you 5 then actual queue length on the disk is about 15.
Can you confirm that VeeamOne logic described by Kirill is correct (collect perfdata only from one node) and please tell when you will plan to fix this - in new release only, or I can open a new one Support Request to get fix on demand?