@flakpyro you probaby had multiple job/schedule on the same schedule I reworked the code and pushed an update and added a doc with the edge cases : https://github.com/vatesfr/xen-orchestra/pull/9635/changes#diff-ef545af2ad06f2759c1a3787f266108b3b2cbc203bc8071bbd847278f3e6a5f0 (will be on doc xo when merged ) fbeauchamp opened this pull request in vatesfr/xen-orchestra open feat(backups): improve incremental replication #9635

Living on the edge at my home lab [image: 1774549062491-screenshot-2026-03-26-at-19-08-50-scrutiny.png] In my other NAS the oldest one is closing up 11 years

And this is the result [image: 1774531386302-screenshot-2026-03-21-at-09-31-45-backup.png]

@Kraken89 I will look into the transition from the previous system to the new one maybe that's the key

@ph7 that is a good news thank you for your patience and help

Great News!!

@florent I rebuilt my XCP hosting environment (everything is faster and bigger stuffed into one rack).... and this issue is now worse. The main changes to this new setup are 2x40Gb networking, faster NFS NVMe NAS, faster pool servers, more memory, much faster CR destination machine with ZFS. Running XCP 8.3 (March 2026 updates) and XO (master a2e33). Replication is leaving many attached to control domain every day with NBD connection set to 2. Changing it to 1 seems to resolve the issue (no more stuck to control domain).

@Pilow We tried that as well, same problem Also tried with a VM on the same newtork, just another VLAN, and we're seeing the same thing. At first we figured it was because one of the xcp-ng's was on a remote site which is connected through an IPSEC VPN, but that wasn't the case.

@Dezerd you just have to start-copy the Replica VM it permits the original job to keep replicating on the VM. there is no failover/failback mecanism AFAIK if you work on replica started VM, you will have to put up in place a replica going to original hosts

png @florent

@MajorP93 said: -disable HA on pool level -disable load balancer plugin -upgrade master -upgrade all other nodes -restart toolstack on master -restart toolstack on all other nodes -live migrate all VMs running on master to other node(s) -reboot master -reboot next node (live migrate all VMs running on that particular node away before doing so) -repeat until all nodes have been rebooted (one node at a time) -re-enable HA on pool level -re-enable load balancer plugin Never had any issues with that. No downtime for none of the VMs. update time again. and same issue I followed these steps -upgrade master -upgrade all other nodes -restart toolstack on master -restart toolstack on all other nodes -live migrate all VMs running on master to other node(s) -reboot master now cant migrate anything else. live migration : NOT_SUPPORTED_DURING_UPGRADE warm migration: fails and VM shuts down immediately and needs to be forced back to life CR backup to another server : NOT_SUPPORTED_DURING_UPGRADE

@olivierlambert done.

@Pilow While that could be true only 1 vm was shuffled to another SR. But all VMs now being backup with CBT enabled. Well... i just check again at backups and after the last scheduled backup at 1am all vms fell back to full backup again.

@florent XO also can't migrate a VM to a new pool with a CD in the drive... It does generate an error, but the error is unclear.

@florent Thanks for the reply. I will rebuild XO and test with current master.

@florent One last update. I reverted to Master branch (6699b) yesterday evening and the backup ran without issues overnight.

502 is an answer coming from your S3, telling the server is having an issue. Adding @florent in the loop