Alright so I got the basically the same situation as described here.
The differences are:
- ESXi 6.7.0 fully patched
- local SSDs and HDDs
The VM is running a common Nextcloud stack (MySQL, Apache2, PHP).
This issue came up just after I upgraded my ESXi to the current patch level (January or February patch I think. There were no other storage related changes on the systems in question. Only thing that comes to my mind is that the VM experiences high disk load, as sometimes both the ghettoVCB and an ssh-based backup overlap.
I can see that the kernel issues messages that point to storage access timeouts. I also just found a log message on the ESXi that reads:
Lost access to volume UUID (name) due to connectivity issues. Recovery attempt is in progress and outcome will be reported shortly.
But there is nothing further reported.
Only solution to get both the VM and ESXi responsible again this is a hard reset of the ESXi host.
I will update this question as my research continues. If anyone has any idea I would highly appreciate your help!