25.11.2025 15:24:33 Succeeded : Preparing to create snapshot
25.11.2025 15:26:05 Succeeded : Failed to inject guest runtime components, failing over to guest agent Connection
25.11.2025 15:27:29 Succeeded : Failed to connect to guest agent, failing over to guest runtime components through PowerShell Direct
25.11.2025 15:28:43 Succeeded : Subscribing to guest processing components (0:00:58)
In general, behavior mods developed for a particular version are not guaranteed to persist through versions. Even just because they are not a part of test plans and if they break, it will go unnoticed.
Same question to me. Not working on v13 latest patch.
We are CSP too. We prefer usage of PSDirect even for our internal stuff and not to open IP communication to VBR server and not to create more proxies inside every subnet.
Already tryed to search for any new registry item via ProcMon.
Moreover. I've found VBR waits for timeout around 5 minutes (sometimes more) between "Preparing to create snapshot" and "Creating VM recovery checkpoint (mode: Veeam application aware processing)" while whole VM backup time is around 10-13 minutes.
So my own calculation:
1. 5 minutes (50% of time) sitting and waiting for timeout for AAIP to fail over via PSDirect
2. 1m 10s for actual disk backup
3. Another 4+ minutes for Hyper-V checkpoints, VMCX/VMRS/VMGS processing, finalizing and etc.
Most of our VMs have minimal HDD data to process - some gigs but it still takes to process 10+ minutes for each VM even with bytes of data (excluding compacting and merging)