Veeam ONE Monitor keeps sending me alerts about VM total disk latency at random intervals.This has been happening for a while now, and only with one specific VM.
If I check datastore latency for this VM in vSphere, it shows read latency spikes up to 250 trillion ms. None of the other VM's have this problem, even the ones on the same datastore.
Some of the spikes in Datastore metrics correlate with spikes in Virtual disk latency metric, eg. the 250T ms datastore read latency spikes correlates with a 45ms Virtual disk read latency spike. At the same time CPU usage is low, disk write was around 80000 KBps.
VM is running CentOS7 64 bit with VMWare tools 10282, SCSI controller is VMware Paravirtual, datastore type is VMFS 6
Server is running HPE Custom ESXi image 6.7
There is a discrepancy with this VM that may or may not be related to this. Under "Summary -> Guest OS" it says CentOS 6 (64-bit), but under "Edit Settings -> VM Options -> General Options -> Guest OS Version" it correctly shows CentOS 7 (64-bit). All the other CentOS7 VMs correctly show the version under summary.
Why is this false metric showing up and how can I get to the root of this issue?
Accepted Solutions
