Sprint
Explorer
- Joined
- Mar 30, 2019
- Messages
- 72
Hi All
So I've been doing a lot of data migration recently as I moved everything off of my 6x4Tb array and squeezed it onto my big array (using the zfs send command), dismantled the old pool, replaced with 8x8Tb drives, rebuilt the pool, then sent the data back (snapshots included). I tried to use the same naming scheme in the hope that the snapshot and replication tasks would pick up where they left off (I had then disabled throughout), but this failed, so I ended up having to restart the replication to my offsite nas by deleting the data at the other end and recreating all the tasks locally.
That had been running for close to 3 weeks, and I expected it to finish today, but when I looked, I noticed that the size of the dataset on the remote end was growing to be bigger than the source dataset?...
I tore it all down and re setup the snapshots and replication but am seeing the same thing, the dataset is 12.46TiB, but the snapshot being replication is 27TiB??? I even deleted every single snapshot and waited for a new one to be generated (for some the reason the replication task won't copy the manually created snapshot i created), and at this point, it'll take months to copy, and would also exhaust my free space at the remote site.
I'm totally stumped, so any suggestions would be greatly received.
(Local Box)
Version: FreeNAS-11.3-U2
VM running in ESXi, 20 CPU cores, 128Gb of ECC DDR4, 2x 8x8Tb RaidZ2 Pools plus some SSD pools, connected to HBAs passed through at hardware level.
(Remote Box)
Version: FreeNAS-11.3-U3.2
Bare Metal Super Mirco board with 10 Core Xeon, 256Gb of Ram, 8x8Tb RaidZ2 pool.
So I've been doing a lot of data migration recently as I moved everything off of my 6x4Tb array and squeezed it onto my big array (using the zfs send command), dismantled the old pool, replaced with 8x8Tb drives, rebuilt the pool, then sent the data back (snapshots included). I tried to use the same naming scheme in the hope that the snapshot and replication tasks would pick up where they left off (I had then disabled throughout), but this failed, so I ended up having to restart the replication to my offsite nas by deleting the data at the other end and recreating all the tasks locally.
That had been running for close to 3 weeks, and I expected it to finish today, but when I looked, I noticed that the size of the dataset on the remote end was growing to be bigger than the source dataset?...
I tore it all down and re setup the snapshots and replication but am seeing the same thing, the dataset is 12.46TiB, but the snapshot being replication is 27TiB??? I even deleted every single snapshot and waited for a new one to be generated (for some the reason the replication task won't copy the manually created snapshot i created), and at this point, it'll take months to copy, and would also exhaust my free space at the remote site.
I'm totally stumped, so any suggestions would be greatly received.
(Local Box)
Version: FreeNAS-11.3-U2
VM running in ESXi, 20 CPU cores, 128Gb of ECC DDR4, 2x 8x8Tb RaidZ2 Pools plus some SSD pools, connected to HBAs passed through at hardware level.
(Remote Box)
Version: FreeNAS-11.3-U3.2
Bare Metal Super Mirco board with 10 Core Xeon, 256Gb of Ram, 8x8Tb RaidZ2 pool.
