Solaris 10 zfs bug

[Mark Felder]

On Jun 3, 2014, at 15:38, Jim Barker <jbarker at nixlike.com> wrote:

> Mark,
> 
> I was looking at the timeframes that the core dumps occurred and around that time we were doing some zfs sends to replicate some data on there.  So either the snapshot creation, zfs send, or snapshot deletion may have triggered the bug.
> 

I personally do quite a few snapshots and zfs send every 15 minutes on a 9.x server and haven't seen this yet. I'd have figured iXSystems would have hit it while doing QA on their TrueNAS product as well if it was as simple as doing a large amount of send/recv testing. However, it's good to have this documented publicly in case someone does come across a similar issue. Perhaps it could be related to zfs attributes on a particular filesystem? (dedup, compression, etc etc)

Thanks for the heads up.