Replication Error no resume after connection loss

TrueNAS General
,
1

Hi guys,

having a probleme here, i dunno how to handle. Google was not my friend coz i couldn’t find a solution.

I do have NAS 0 and NAS 1. NAS 0 is replicating its snapshots to the offsite NAS 1 via an replication task. I used the guided replication task helper and its a Scale to Scale connection via ssh. Its doing perfectly fine as long as there is no interupption with the connection between both NAS’es

My problem is, that the sync isn’t resuming after an connection loss (24h reconnect in germany) happens. And the Snapshots i wanna push are to large to handle in 24h, because the connection speed is very very slow.

its giving me the following error message:

Replication “NVMEPool - Backup 5TB/BackupNVMEPool” failed: Last full ZFS replication failed to transfer all the children of the snapshot NVMEPool@auto-2025-04-04_02-15. The snapshot Backup Gameserver 8TB/TruenasNVMEPool/.system@auto-2025-04-04_02-15 was not transferred. Please run zfs destroy -r Backup Gameserver 8TB/TruenasNVMEPool@auto-2025-04-04_02-15 on the target system and run replication again…

Can anybody help me to enable/find/configure the resume feature or give me that little hint what checkboxes i have to click to get it going?

Thx :slight_smile:

2

Afaik, before restart, you should follow the tip and destroy the snapshot indicated.
It randomly happens to me too (i use a Powerline and Is not so reliable), and i do that to resume correctly

3

I get this message every time when the connection is lost. The snapshot is to big to be sent in less than 24h. Result is in no Backup at all. That’s why I want to resume the replication.

4

Correct me if i misurunderstand… You have those disconnection not random but systematic every 24h, and this time Is not enough to transfer data, right?

5

Yeah that’s right

6

Honestly i don’t know if there Is this possibility via GUI.
But according to this old post should be possible at least manually:

Also sound strange to me that has not been implemented yet… Hope someone more Expert can point you in the right direction