We have two VMs with basic 50GB Debian OS hard disks
These Vms share a 500GB Thick Provision Eager Zeroed disk and both Vms have a mount point sbd1/var/mydata
We use OCFS2 as a clustered file system and both VMs can see and access the 500GB drive. Originally the 500GB drive was created when setting up the first VM, it was zeroed in and when the second VM was built this disk was added.
Both VMs can access the 500GB disk independently, when each other is running or shutdown.
The issue I have is if I create and write a 2GB file to sbd1/var/mydata on VM1 it takes 8 seconds, if I do the same on VM2 it takes 25 seconds. Is it possible that VM2 has a more difficult path to write to the 500GB drive that was originally created when building VM1
I ask as we have had issues with OCFS2 clustering services throwing errors when writing large amounts of data to the shared 500GB drive
Any ideas
Cheers
