SOLVED zfs send/receive not starting

[RotsiserMho]

I'm at a loss. I'm trying to migrate data from an old FreeNAS server running 9.2.1.7 to a new one running 11.3-U4.1. The new server has a pool called "data" which is a RAIDZ2 across 6 8TB drives. I migrated one pool by installing the drives in the new server and doing a replication via the GUI. So far, so good. But I have another pool on the old server that is a RAIDZ1 across 3 2TB drives that I can't install on the new server for lack of SATA ports. So I'm trying to migrate it via zfs send/receive over ssh. Here's what I tried:

1. Create snapshot of old pool (called Important): zfs snap -r Important@Backup1
2. Send to new server: zfs send -Rv Important@Backup1 | pv | ssh root@[new server] zfs receive -v data/Important

There are no errors, but the following line just prints to the console every second. I'm not sure what's happening, but neither computer has any significant CPU load.

23:16:01 174K Important@Backup1
23:16:02 174K Important@Backup1
23:16:03 174K Important@Backup1
23:16:04 174K Important@Backup1

I don't think it's working correctly.

 

[Apollo]

Try with extra v's into the cmd:

zfs send -R -vv Important@Backup1 | ssh root@[new server] zfs receive -vv data/Important

I am unaware of the | pv commad so I won't comment on it.

If I recall things properly, you would get the repeating "174K Important@Backup1 " message if you don't have proper ssh authentication done.

See if you can run:
ssh root@[new server] zpool stauts or other commands. The intent is to verify you can actually run command remotely.

 

[RotsiserMho]

Ah, yes, the extra v's trick. The pv command allows for reporting progress.

You were absolutely right, I was using password authentication and I think it got stuck waiting for password entry in the middle of the pipeline or something. Switching to key-based authentication did the trick and it's transferring fine now. Thanks!

 

[Apollo]

Ah, yes, the extra v's trick. The pv command allows for reporting progress.

You were absolutely right, I was using password authentication and I think it got stuck waiting for password entry in the middle of the pipeline or something. Switching to key-based authentication did the trick and it's transferring fine now. Thanks!

Thank you for providing feedbacks on your issue. I am running my replication process using a Python script, unfortunately, if I try to use the "|pv" option, I am unable to display real time status of the replication.

It does work as a non script based command.
One thing I have noticed its that it just provides information on how long it took to send that much data. I believe there is an option that should indicate the overall transfer state, instead. I have seen it, but don't know the command.