All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH] Relax exit codes for policy-driven lvconvert --repair.
@ 2010-01-07 20:02 Petr Rockai
  2010-01-07 21:00 ` Petr Rockai
  0 siblings, 1 reply; 2+ messages in thread
From: Petr Rockai @ 2010-01-07 20:02 UTC (permalink / raw)
  To: lvm-devel

Hi,

due to the interpretation of "replace" policies as "replace if possible,
downconvert otherwise", we should treat successful downconversion as a
success in these cases, even if we were unable to replace the missing
devices with new ones.

The attached patch also makes the "nothing to repair" condition a
non-error (with and without --use-policies... it's arguably not really
an error). This would hopefully address RHBZ 552723, if it is indeed the
case (as it seems) that the problem there is that the missing devices
have already returned by the time that lvconvert --repair kicks in (and
therefore it does not detect anything to repair).

Yours,
   Petr.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: lvconvert-relax-failures-with-policy.diff
Type: text/x-diff
Size: 5086 bytes
Desc: not available
URL: <http://listman.redhat.com/archives/lvm-devel/attachments/20100107/4728a608/attachment.bin>

^ permalink raw reply	[flat|nested] 2+ messages in thread

* [PATCH] Relax exit codes for policy-driven lvconvert --repair.
  2010-01-07 20:02 [PATCH] Relax exit codes for policy-driven lvconvert --repair Petr Rockai
@ 2010-01-07 21:00 ` Petr Rockai
  0 siblings, 0 replies; 2+ messages in thread
From: Petr Rockai @ 2010-01-07 21:00 UTC (permalink / raw)
  To: lvm-devel

Hi again,

Petr Rockai <prockai@redhat.com> writes:
> due to the interpretation of "replace" policies as "replace if possible,
> downconvert otherwise", we should treat successful downconversion as a
> success in these cases, even if we were unable to replace the missing
> devices with new ones.
>
> The attached patch also makes the "nothing to repair" condition a
> non-error (with and without --use-policies... it's arguably not really
> an error). This would hopefully address RHBZ 552723, if it is indeed the
> case (as it seems) that the problem there is that the missing devices
> have already returned by the time that lvconvert --repair kicks in (and
> therefore it does not detect anything to repair).

an amended patch that also ensures that MISSING_PV removal is done even
if device replacement fails is attached, and an interdiff with only the
new changes.

Yours,
   Petr.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: lvconvert-dont-skip-remove_missing_empty_pv.diff
Type: text/x-diff
Size: 1741 bytes
Desc: not available
URL: <http://listman.redhat.com/archives/lvm-devel/attachments/20100107/40fc3df7/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: lvconvert-relax-failures-with-policy-2.diff
Type: text/x-diff
Size: 6577 bytes
Desc: not available
URL: <http://listman.redhat.com/archives/lvm-devel/attachments/20100107/40fc3df7/attachment-0001.bin>

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2010-01-07 21:00 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2010-01-07 20:02 [PATCH] Relax exit codes for policy-driven lvconvert --repair Petr Rockai
2010-01-07 21:00 ` Petr Rockai

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.