The environment is a HUSVM with luns mapped to a UCS blade running RedHat.
We needed to upgrade some of the cluster packages a Redhat cluster so we took a TI snapshot of the production boot volume in case we needed to restore.
Once the upgrade was complete they verified that the changes had been made to the packages. Later that weekend the server was rebooted a the admin noticed that the changes he made were no longer applied.
The TI snapshot was never removed at this point and is still there, but we just assumed that there was a problem with the package install procedure.
We proceeded to upgrade the packages again only this time we decided to clone the volume and mount it to a new system to upgrade from there. I created a SI clone of the OS volume and mounted it to a new test host. When the admin checked the packages on the clone they noticed that the clone contained the upgraded packages from the first time.
We verified on the P-Vol that the packages did not have the updates applied but can see that the clone volume does.
Apparently the host is running off of the snapshot now and we have no idea why. The original P-Vol has always been allocated to this host and I have validated that everything is mapped correctly.
Has anyone seen this behavior before. Would removing the snapshot be a problem for the host?