Rebooting Virtual Machine’s on an ESXi 5.1 cluster is a regular part of maintenance done on a monthly basis for us. We turned off one of the VMs and it refused to turn back on giving us error:
“A general system error occurred: Unknown Error.“
“Cannot create file /var/run/vmware/tickets/vmtck-XXXXX for process hostd-worker because the inode table of its ramdisk (root) is full.”
Rebooting the host in the cluster may have resolved this issue quickly, but we had some time and wanted to find a root cause for this error.
The issue is with the inode table being full we can see this by running the command: esxcli system visorfs ramdisk list
It turns out that the monitoring software (snmp) implemented filled up the space. Ooops!
VMWare’s Official Fix:
After following the KB article we were able to turn the VM on with no problems!