[illumos-Developer] Replay problems with new zvol_replay_truncate (was iSCSI UNMAP...)
George Wilson
gwilson at zfsmail.com
Tue Mar 1 07:41:51 PST 2011
Dan,
What you're seeing makes sense since the breakpoint is after the log
record is created in-core but before the zil_commit() which will sync it
to disk. Now set a breakpoint after zil_commit() but before the
txg_wait_synced() and let's see what happens (make sure you're zvol is
big enough so that there's plenty of work to do).
- George
On 2/28/11 1:16 PM, Dan McDonald wrote:
> On Fri, Feb 25, 2011 at 08:57:51AM -0800, George Wilson wrote:
>
> <SNIP!>
>
>> We also need to test the log/replay logic. So I would recommend
>> setting a breakpoint after zvol_log_truncate() and forcing a panic.
> I did just this, and forced a panic. (Sidenote: once the dump finishes, the
> machine locks after printing "100%" and I get no actual dump...).
>
> After the panic, I booted in with kmdb -d, and set this breakpoint
> immediately:
>
> ::bp zfs`zvol_replay_truncate
>
> figuring the system would break there, since I wrote my unmap to the log with
> zvol_log_truncate().
>
> No luck --> machine boots all the way to multiuser and login. And the blocks
> are not reclaimed. Only a low-level format from the iSCSI initiator does
> anything. :-/
>
> I tried one change prior to sending this note, and that was to assign
> itx_private to the zvol state in zvol_log_truncate() like the
> zvol_log_write() call does. Also no luck.
>
> I've updated the webrev:
>
> http://www.kebe.com/~danmcd/webrevs/unmap/
>
> but I'm sure I'm missing something in zvol_log_truncate(). I just don't know
> what precisely.
>
> I'm unavailable the rest of the day today, but I will be at least reading
> mail v. late this evening, and maybe even getting some more work done if I
> have a flash of insight during my unavailable time, or someone else does.
>
> Thanks again for your time& patience,
> Dan
>
> _______________________________________________
> Developer mailing list
> Developer at lists.illumos.org
> http://lists.illumos.org/m/listinfo/developer
More information about the Developer
mailing list