One day recently we discovered at Inksters that we had, on the face of it, lost all work created from 11.30am the previous day. It transpired that a Systems Engineer at the data centre was performing a routine volume extension the night before which unfortunately corrupted the volume. The Engineer promoted a replica of this volume on the backup group but unfortunately this replica was also corrupt. The end result of all this was our data being backed up to what it looked like at 11.30am. The data centre did not notify our supplier of this issue and we discovered it ourselves. It took most of the day for the issues to be resolved and the missing documents located (they were saved somewhere in that data centre) and restored to our cloud server. This caused us a lot of disruption with some documents having to be recreated by us as we did not know how long it was going to take to resolve the issue.
I have been assured that procedures, error and incident reporting and disaster recovery procedural changes are being put in place to prevent this from happening again.
This incident has certainly not put me off the cloud as the data was still there and was reconstituted. If anything it demonstrated that if something goes wrong the backup system is such that your data can be easily restored.
The supplier and the data centre were excellent in their response, concern and reassurances. Indeed the management of the data centre even invited me on a tour of the data centre and my next post on The Time Blawg will be about that visit.