dm-devel.redhat.com archive mirror
 help / color / mirror / Atom feed
* dm-cache suspend semantics
@ 2015-08-19 11:26 Thanos Makatos
  2015-08-19 18:42 ` Alasdair G Kergon
  0 siblings, 1 reply; 2+ messages in thread
From: Thanos Makatos @ 2015-08-19 11:26 UTC (permalink / raw)
  To: device-mapper development; +Cc: heinzm, thornber, Mike Snitzer

I'd like to know whether it is possible for a cache target for which
there are dirty blocks to continue writing back dirty data after it
has been suspended. Can this happen? I checked the code and
cache_postsuspend() suggests that after the target is suspended no
write back I/O will be initiated (I also did a quick test that
confirms this), however I'd appreciate an answer from the maintainers
as I might have misunderstood.

-- 
Thanos Makatos

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

* Re: dm-cache suspend semantics
  2015-08-19 11:26 dm-cache suspend semantics Thanos Makatos
@ 2015-08-19 18:42 ` Alasdair G Kergon
  0 siblings, 0 replies; 2+ messages in thread
From: Alasdair G Kergon @ 2015-08-19 18:42 UTC (permalink / raw)
  To: Thanos Makatos; +Cc: heinzm, device-mapper development, thornber, Mike Snitzer

On Wed, Aug 19, 2015 at 12:26:20PM +0100, Thanos Makatos wrote:
> I'd like to know whether it is possible for a cache target for which
> there are dirty blocks to continue writing back dirty data after it
> has been suspended. Can this happen?

A suspended device-mapper device must not generate any I/O to the
devices underneath it while it is suspended.

Alasdair

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

end of thread, other threads:[~2015-08-19 18:42 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2015-08-19 11:26 dm-cache suspend semantics Thanos Makatos
2015-08-19 18:42 ` Alasdair G Kergon

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).