All of lore.kernel.org
 help / color / mirror / Atom feed
* librbd diff_iterate question
@ 2016-09-14 19:13 Max Yehorov
  2016-09-14 20:10 ` Jason Dillaman
  0 siblings, 1 reply; 2+ messages in thread
From: Max Yehorov @ 2016-09-14 19:13 UTC (permalink / raw)
  To: ceph-devel

Hi,

according to code comments: diff_iterate2 returns the differences
between two versions of an image via a callback, which gets the offset
and length and a flag indicating whether the extent exists.

Now, there are two code paths to calculate diff, one that uses
fast-diff/obj-map and the other which asks osds. It appears that only
fast-diff code path can set "exists" flag to false in the
diff_iterate's callback. Is that the correct assumtion?

thanks!

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

* Re: librbd diff_iterate question
  2016-09-14 19:13 librbd diff_iterate question Max Yehorov
@ 2016-09-14 20:10 ` Jason Dillaman
  0 siblings, 0 replies; 2+ messages in thread
From: Jason Dillaman @ 2016-09-14 20:10 UTC (permalink / raw)
  To: Max Yehorov; +Cc: ceph-devel

On Wed, Sep 14, 2016 at 3:13 PM, Max Yehorov <myehorov@skytap.com> wrote:
> It appears that only
> fast-diff code path can set "exists" flag to false in the
> diff_iterate's callback. Is that the correct assumtion?

The fast-diff code path within diff_iterate can set the "exists" flag
to true (1) if the object exists and has been updated since the "from"
snapshot (i.e. there is a delta), otherwise it is set to false (0).

-- 
Jason

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

end of thread, other threads:[~2016-09-14 20:10 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2016-09-14 19:13 librbd diff_iterate question Max Yehorov
2016-09-14 20:10 ` Jason Dillaman

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.