The instability continues with any kernel we operate, at this point of time we assume the issue to be in the hardware layer (with the RAM).

A replacement ordered and will be delivered by Friday (at most!) -- and all servcies will be transferred to the new server at that time (or earlier!).

Host-node is running on,

5.9.9-1.el7.elrepo.x86_64

which is the latest available from elrepo for el7/centos7, we hope this will further help with the stability.

Hello,

We're updating the latest Kernel now and rebooting the host server shortly.

Thank you.

Kernel installed, reboot in-progress..

Hello,

We're updating the Kernel with 5.XX now and rebooting the host server shortly.

Thanks

we are debugging this host-node further, we assume it could be a swap-related issue, the node has more than enough ram but swap area issues could be causing stability issues. disabling swap for now.

host-node is back online with the 4.xx kernel, we continue to monitor the node now. VMs booting up one by one again.

4.xx kernel-lt installed, host-node under reboot now

We notice the instability with the newer kernel as well, we're switching to 4.xx kernel-lt ELRepo based kernel now.

latest kernel installed, host-node has been rebooted now. VMs to boot up one by one now

Due to instability in the current kernel with KVM based VMs, we're updating to a newer kernel and rebooting the host-node to resolve this issue.

Updates to follow.