How to maintain periodic snapshot, replication tasks, etc. through a pool rebuild?

[redoak42]

We're rebuilding our two freenas systems to be raidz2 rather than raidz1 (FreeNAS Mini hardware). When I rebuilt the back up system, and exported the pool, several items such as periodic snapshots seem to have went away with the pool. While I can write everything down and rebuild the configs by hand, I'd much prefer to not have to do that. I'm using ZFS send/receive to move everything around as I rebuild the pools, so the paths, snapshots, etc. should all be identical when I'm done to before I started the rebuild, just the underlying pool will be a raidz2 on bigger disks.

Before I change the primary system, I'm going to temporarily move the system data set to the boot drive, and then take a backup of the config. If I restore the config after I have the pool and file systems back in place, will that get me back the tasks, etc. or am I way off base here?

Thanks!

 

[Ericloewe]

If all the paths are the same, you might get away with pulling a switcheroo while FreeNAS isn't watching (i.e. the server is off).

 

[redoak42]

If I see what you're getting at, I could sync the data from the first to second system so the second system has disks identical to the first (but with bigger disks), power down the both systems, and switch disks.

Do you think the config restore would also work, or is that data I'm losing held somewhere else (such as the .system file system), which I don't believe would get restored with the config?

The configs are not that complex, so by hand wouldn't be the end of the world.

 

[redoak42]

Searching around the DB file, it seems those settings are in there, so a restore after the rebuild should bring them back. Thanks for switcheroo possibility. That had been one of my plans for DR depending on what component failed, but given the parts I have on hand (drives aren't coming in as fast as I would like) and a bit of time hustle to rebuild the backup array that had failed due to a disk being kicked out while another was rebuilding, it wasn't so easy.

I've seen on the couple of minis that I have what appear to be perfectly healthy WD Red disks (cksums, SMART) that then operate fine afterwards get "removed" on occasion, and this occasion was very unfortunate. My own fault for having single parity, but that's getting fixed I immediately rebuilt the backup array with DP and 6TB disks, but that gives us very little breathing room, so the primary array is getting 8tb disks from the get go, and then I'll rebuild the backup array again, or more likely swap one disk at a time and expand.