Spurious 'replication failed' emails

[rogerh]

Since the last three updates every successful snapshot replication generates a 'replication failed' email warning sent to my account. Since I am doing hourly snapshots on five datasets this is more than a little annoying. I have reported it:

https://bugs.freenas.org/issues/12143

Anyone else getting this?

 

[tmueko]

Same Problem here:

Replication failed at daten/jails/acc-2016@auto-20151027.1200-4w -> auto-20151027.1500-4w

Hello,
The replication failed for the local ZFS daten/jails/acc-2016 while attempting to
apply incremental send of snapshot auto-20151027.1200-4w -> auto-20151027.1500-4w to 192.168.222.20

Next Time:

Replication failed at daten/jails/acc-2016@auto-20151027.1500-4w -> auto-20151027.1800-4w

Hello,
The replication failed for the local ZFS daten/jails/acc-2016 while attempting to
apply incremental send of snapshot auto-20151027.1500-4w -> auto-20151027.1800-4w to 192.168.222.20

 

[rogerh]

Same Problem here:

Replication failed at daten/jails/acc-2016@auto-20151027.1200-4w -> auto-20151027.1500-4w

Hello,
The replication failed for the local ZFS daten/jails/acc-2016 while attempting to
apply incremental send of snapshot auto-20151027.1200-4w -> auto-20151027.1500-4w to 192.168.222.20

Next Time:

Replication failed at daten/jails/acc-2016@auto-20151027.1500-4w -> auto-20151027.1800-4w

Hello,
The replication failed for the local ZFS daten/jails/acc-2016 while attempting to
apply incremental send of snapshot auto-20151027.1500-4w -> auto-20151027.1800-4w to 192.168.222.20

Are these allegedly failed snapshots actually being received on the target machine? In my case they are.

 

[depasseg]

It's happening to me too.

 

[tmueko]

In my case the snapshots are in deed on the destination maschine.

 

[tmueko]

I check the snapshots: In my case the error message is totally correct: the two snapshots defer!:

[root@baal5] ~# zfs send daten/wolfgang/jails/acc@auto-20151112.1500-4w | zstreamdump
BEGIN record
hdrtype = 1
features = 4
magic = 2f5bacbac
creation_time = 56449b63
type = 2
flags = 0x0
toguid = c04620d45b343858
fromguid = 0
toname = daten/wolfgang/jails/acc@auto-20151112.1500-4w
END checksum = bbfa98c5c9b61547/44550183f6803d3b/4ca97283a36da380/643ddbba99f85c90

[root@wolfgang] ~# zfs send daten/jails/acc@auto-20151112.1500-4w | zstreamdump
BEGIN record
hdrtype = 1
features = 4
magic = 2f5bacbac
creation_time = 56449b63
type = 2
flags = 0x0
toguid = c04620d45b343858
fromguid = 0
toname = daten/jails/acc@auto-20151112.1500-4w
END checksum = bbfa96324af14106/412f2339ba0bc80f/d3e48ebef9457770/25c90fb799523980

I will delete the backup and have it rebuilded.

 

[saurav]

I have this too. This definitely started for me since I upgraded to 201511040813, and wasn't happening till then. In my case, this happens only for one specific jail which is recursively snapshotted and replicated to the backup among others. The backup seems to receive all snapshots properly and "zfs diff" on failed replication snapshots seems identical on both sides. Also, I get the replication failed message between each snapshot range only once, which probably means the replication actually succeeds and something else fails after replication.

The only issue seems to be that the old snapshots from the backup are not cleared automatically. I manually cleared all the stale snapshots on backup, but it didn't help. The next replication still failed.