SOLVED Replication stops with sshd fatal error

[Toddebner]

I have run into a replication problem where the replication stops at 99% and the source machine throws the following error.

FREENAS sshd[###]... fatal:readfrom socket failed....

The source machine is at my office and is running 9.2.9
The receiving machine is at my home and is running 9.2.9

I am connecting using dynamic dns and a host name as I dont' have a static ip at home.
Also, I am not using a site to site vpn.

I took the receiving machine into my office. completed a replication locally.
Then reset the ip/host name info in the zfs replication setup
Took the receiving machine home
Started it up
Ran a snapshot at the source machine
Replication started up and completed
The next night zfs replication started up at appropriate time (2200, after a server backup had completed to the freenas source machine)
It ran until hitting 99% complete. then threw the error

I have tried this twice now with the same results.

Any ideas on what to do here ?

Both boxes are running supermicro x7sbe motherboards, intel q6600, 8GB ECC Ram, 4-2T drives

 

[dlavigne]

Did you ever figure this out?

 

[Toddebner]

Never found a solution written up anywhere.

So, I recopied the public key from the push box and put it on the "root" user in the pull box. Rebooted both machines and it started back up.

Not sure if it was the public keys or the reboots, but its all working now.

:)