ReadyData Replicate failing with “zfs: Bad file number (2070)” 5200

By | May 19, 2014

Are you seeing this error on your replicates

[2014-05-14 00:01:58] [INFO] Connect to destination
[2014-05-14 00:01:58] [INFO] Connected
[2014-05-14 00:01:58] [INFO] START block sent
[2014-05-14 00:01:59] [INFO] START block received
[2014-05-14 00:01:59] [ERROR] Target unable to resume
[2014-05-14 00:01:59] [ERROR] Transfer failed
[2014-05-14 00:01:59] [INFO] Sent total 83.08 KB
[2014-05-14 00:01:59] [INFO] Average speed 41.54 KB/s
[2014-05-14 00:02:20] [ERROR] zfs: Bad file number (2070)
[2014-05-14 00:02:20] [INFO] Disconnect
[2014-05-14 00:02:20] [INFO] Retry failed
[2014-05-14 00:02:20] [ERROR] Job failed

This is due to a problem with snapsots that are used for replication, if you have SSH access try viewing the replicate log by entering the following

tail -f /opt/replicate/var/log/*.log

Then start the replicate job.  You will probably see this error on the destination filer

[ERROR] zfs: destination has snapshots

To resolve this issue access the destination filers GUI.  Access the Shares menu then right click the LUN and press “Import Snapshots”.

Import-Snapshot

This should sync the snapshots.

Now repeat the process on the source filer to import the snapshots.

Once this is complete retry the replicate job and it should start transferring.

It will resync the entire volume so be patient and check the logs for details.  If at the end of this it still fails contact Netgear Support.

Contact Netgear support before attempting any thing first to advise of your intentions.