We recently migrated our Commserve to a new physical server and now several weeks post migration we are getting errors on about 90% of our synthetic fulls when they reach the end of the job, they will continously run at 99% but looking at the synth full logs we see alot of this:
30236 5280 06/05 12:34:18 13071395 SdtDecryption::ProcessBuffer:218 Cannot decrypt data for afid [28434509]. cur_afid [28725857], orig_afid [28434509]
30236 574c 06/05 12:34:18 13071395 SdtDecryption::ProcessBuffer:218 Cannot decrypt data for afid [28434509]. cur_afid [28725857], orig_afid [28434509]
The event viewer gives the generic Unable to communicate with Media Agent or Pipeline error. Opened a high priority ticket about a month ago and we’re being told we need to seal every DDB store in our environment and re-base. We have several thousands of VMs with some over 60TB in size. I feel like re-basing is going to take weeks, and we may not even have enough storage to do it.
Any other thoughts or ideas would be VERY much appreciated. We had to re-install the software a few times on the new box - They suspect that somehow when the CSDB was moved from the old physical hardware to the new that somehow jobs were written in-between this time that are causing issues but we were very careful during the migration, I selected new DB and used the CSDB Recovery tool after with someone from support on the line the whole time and making sure the services on the old hardware were never started at the same time.
The only other issue we had was the Replace Media Agent tool failed on some mount paths, but I assume we’d be looking at different issues in the logs.
