Yesterday, I had a ESXi server working. Total memory required by the deployed virtual machines is higher than the physical RAM. The disk also has a few Gigabytes of free capacity.
Today my server stopped normal operation! According the to the performance reports, memory usage exceeded the physical memory and the disk does not have enough free space to swap it.
In 'Configuration > Storage' , the datastore is disabled and is not accessible. SSH connection to ESXi is possible, however, any disk command takes a long time to execute (e.g. df). I tried to halt some VMs to free up some memory. But the 'Power Off' commands stalled at 95% in Recent Task window. I looks the disk faced to critical problem.
Is there anyway to bring the system to normal state without reboot/halt?
Any help is highly appreciated. Thank you so much.
I would not expect to be able to resolve this without some downtime. Here, you really should check the health of your local disks.
If you have important data or systems running, they're probably not running optimally with your storage subsystem issue. Your best clues as to what happened are the disabled datastore and to check the "Alarms" tab for the host using the vSphere client.