SOLVED FreeNAS 11 RC4 zfs send receive causing core dump

[Mr Steven]

I also had the issue in 11 RC3 - I'm pretty sure this is a bug or maybe someone can explain what I am doing wrong.

I want to copy a dataset to another volume in the same system. So I do...

# zfs snapshot -r main/archives@backup
# zfs send -R main/archives@backup | zfs receive -vF vault/archives

The snapshot is created but the send receive operation eventually throws a (zfs), uid 0: exited on signal 6 (core dumped)

You can see the dataset gets created in the Volume manager. You can see it on the CLI /mnt/vault/archives but you can't see any data in there unless you reboot the server.

 

[Ericloewe]

Hardware? Copy of the stack trace?

 

[Mr Steven]

I'm not a dev and I have no idea how to do a stack trace so please excuse my ignorance. Is this it?

Core was generated by `zfs receive -vF vault/archives'.
Program terminated with signal SIGABRT, Aborted.

Thread 1 (LWP 102788):
#0 0x0000000801376d0a in ?? ()
#1 0x0000000801376cd4 in ?? ()
#2 0x0000000000019184 in ?? ()
#3 0x2791453cb8a1acac in ?? ()
#4 0x00007fffffff7334 in ?? ()
#5 0x0000000802e40000 in ?? ()
#6 0x00007fffffff7350 in ?? ()
#7 0x0000000801376c49 in ?? ()
#8 0x2791453cb8a1acac in ?? ()
#9 0x0000000000000021 in ?? ()
#10 0xffffffdf00000021 in ?? ()
#11 0xffffffffffffffff in ?? ()
#12 0x00000008ffffffff in ?? ()
#13 0x0000000000000016 in ?? ()
#14 0x00007fffffff7450 in ?? ()
#15 0x0000000801083143 in ?? ()
#16 0x00007fffffff7338 in ?? ()
#17 0x00007fffffff73e0 in ?? ()
#18 0x0000000000000018 in ?? ()
#19 0x00007fffffff7c70 in ?? ()
#20 0x00007fffffff72e8 in ?? ()
#21 0x8080808080808080 in ?? ()
#22 0x000000080063037f in ?? ()
#23 0x00000000000016f8 in ?? ()
#24 0x000000000000037f in ?? ()
#25 0x0000000000000000 in ?? ()

Hardware is a HP N40L Microserver.


Update:

I set up a VMware test machine and the same commands work without any issue. The datasets are different though.

 

[Mr Steven]

Well after some further testing it looks like Corral may have come back to bite me on the backside - again.

I set up two new disks. Created a new dataset on one of them and then migrated my archives dataset to it via rsync.

I ran the zfs snapshot send and receive commands on these newly created disks and it worked.

So it seems Corral did something bad to zfs? Is there a way of fixing this?

 

[Ericloewe]

Yes, follow the instructions available in the Resources section to set the correct aclmode.

 

[Mr Steven]

As per the instructions in the resources section I ran:

# zfs set aclmode=restricted main/archives

And that worked, thanks!