ZFS Write Lockup

[Daniel Staal]

--As of October 5, 2011 6:37:17 PM -0400, Dave Cundiff is alleged to have 
said:

> Its for a backup service I've been working on. It takes a snapshot
> hourly of all 17 zvols. I was planning on keeping them for a month.
>
> I had the same thought about the snapshots and deleted them all
> yesterday. It appears there is some issue with keeping that many. I
> removed them all and the zvols are now functioning correctly. Its
> strange that the large number didn't cause incremental slowdown. While
> the snapshots were still there the IO was normal when it wasn't acting
> up. Just it would have spurts of almost total lockup until I performed
> a snapshot removal operation or 2.

--As for the rest, it is mine.

Look at sysutils/zfs-periodic.  It does that type of backup, although it 
doesn't keep quite as many.  (Several hourly, daily, weekly, and monthly. 
Configurable.  I have 38 filesystems, and 780 snapshots. (Only one volume, 
though.))

And that definitely sounds like a bug.  Gradual slowdowns would be expected 
if you were just reaching performance limits, but sudden stops sound like 
there's a condition that doesn't work someplace, or some loop the system 
can get into.

Daniel T. Staal

---------------------------------------------------------------
This email copyright the author.  Unless otherwise noted, you
are expressly allowed to retransmit, quote, or otherwise use
the contents for non-commercial purposes.  This copyright will
expire 5 years after the author's death, or in 30 years,
whichever is longer, unless such a period is in excess of
local copyright law.
---------------------------------------------------------------