Previous full backup chain transformation failed Error: Agent: Failed to process method {Transform.Patch}

[lando_uk]

"Previous full backup chain transformation failed Error: Agent: Failed to process method {Transform.Patch}: Application is shutting down."

Some synthetic full's didn't transform at the weekend. Got this error on several jobs across several repositories. Which if it was an IO issue is strange that it happened across more than one.

Logs uploaded. but whilst i wait for support (03822462.) any ideas?

[Egor Yakovlev]

Hi Mark!

It's hard to tell without logs inspection. I can see your case was picked up by Veeam Support Engineer and under investigation.
Hold on tight and wait for it, please!

/Egor

p.s. Congratulations on 6 years with Veeam Forums!

[lando_uk]

OK i'll wait... just impatient

I've done some test restores and the chain seems to be ok. Just no transform, with no reason given. It'll be a pain if i have to run some new Fulls...

6 years huh, not that many support calls opened in that time, kudos to you lot over there.

[lando_uk]

So i've had a response and i've change the HangedAgentKillTimeout reg on my servers.
What i don't get is, what would cause the Transport engine to terminate different jobs, on 3 different repositories in the time interval. I can understand one job failing, but lots of them? And after all this time? There must be some external reason, maybe the Veeam database hung for a while or something? Would merge tasks reference the SQL DB constantly and fail if there was a blip? (even though i cant see any blips...)

[Egor Yakovlev]

First, let's confirm that provided solution worked - I guess next merge is coming on a following weekend, right?
Then we can think of a possible next step - it is hard to find a root cause of a "one-time" timeout when job does it with no visible reason - it might have been anything, from "[insert any reason here]" down to automatically installed windows\sql updates \ antivirus scan \ human factor \ latency spike \ network glitch? However, if transform will fail again after suggested fix, it will signal us for a repeating problem which will be easier to track down.

/Thanks!

[lando_uk]

Hi. The transform is working this week, but its very slow due to having to process 10 increments rather than the normal 5. It'll probably leak over into tomorrow on certain jobs.

(...one day i'll get new hardware with ReFS volumes)