Big question, let me lay some groundwork. We have two TrueNAS systems, one at the operations center, the other at the DR site. The operations center site is set up to replicate 4 pools to the DR site unit. Works great. But the last time we performed the OS update to TrueNAS-13.0-U6.3 on the two units, the update must have busted the replication, because they went completely offline from each other. Could not get the replicas to reconnect, and had to destroy the replicas on the DR site unit, and rebuild the whole mess from scratch, and resync from ground zero. OUCH-much time spent rebuilding the wheel! Pertinent information was that the upgrades were done on the two units within 1 minute of each other, as we triggered them simultaneously, hoping that they would upgrade and sync back up seamlessly, and they didn’t. What are the rules for upgrading to the next STABLE release when you have this type of replication in place?
I could be reading this wrong, but how can an OS update “unsync” your datasets?
The update should not (and does not) touch your storage pools.
Are you using some sort of HA setup? The way you’re using the word “sync” implies it’s not a typical one-way ZFS replication.
The operations center is set up under Tasks\Replication Tasks to replicate 4 volumes in pool0 to the identical IxSystems Truenas hardware unit at the DR site. When we performed that U6.3 update, it broke the replication tasks, it messed up the destination pool, and on the DR site’s unit, it even fouled up the Sharing/iSCSI setup on the unit to the point that I had to rebuild all of the targets and extents and their bindings to even get the disks back online. We spent a couple days fixing and getting it back operational, so you can see our hesitancy to install 13.0-U6.7. Beings that we’re using the stock-standard replication tasks, is that considered a HA environment? We didn’t think so, but we’re always open to re-education. “Sync” was a loose word, I should have said “replication” in this context. Is there a way to “pause” replication, perform the U6.7 upgrade, and then re-enable or “un-pause” replication for it to pick back up and continue normally, without it trashing it back into oblivion like it did before??