Hello Daniel,

I recommend that you use the same link to talk to croit:

https://www.croit.io/contact

Caveat: it goes to sales. As far as I know, croit has a production-ready
patch for Ceph-18.2.4 and submitted most (all?) of their patches upstream,
but there is no ready-made backport to your Ceph version (19.2.x). Sales
guys might view this as an opportunity to sell you developer time necessary
for backporting and/or try to sell a support contract.

Or, if you don't want to talk to croit, the main patch against taking the
wrong directory snapshot when mirroring is this one:

https://github.com/ceph/ceph/pull/62281

Note that this does not contain the functionality needed to automatically
clean up after bugs in old versions of cephfs-mirror. A patch provided by
croit contains this functionality, but I could not find any equivalent in
public pull requests (probably didn't search well enough).

You also may want to check if your Ceph build contains this:

https://github.com/ceph/ceph/pull/61100


On Wed, Apr 16, 2025 at 11:26 PM Daniel Schreiber <
daniel.schrei...@hrz.tu-chemnitz.de> wrote:

> Hi Alexander,
>
> we are observing similar behaviour on two 19.2.1 clusters that are not
> in production yet. If you want to debug something, please let me know.
>
> We saw the following bugs:
>
> * missing data on the destination
> * data from the past on the destination
>
> Regards,
>
> Daniel
>
> Am 21.03.25 um 06:38 schrieb Alexander Patrakov:
> > Hello Vladimir,
> >
> > Please contact croit via https://www.croit.io/contact for unofficial
> > (not yet fully reviewed) patches and mention me. They are currently
> > working on a similar problem and have solved at least the following:
> >
> > * Crashes when two directories are being mirrored in parallel
> > * Data corruption (snapshots on the destination having missing files
> > or files "from the future") due to incorrect cache maintenance in
> > libcephfs
> > * Confusion that results in the failed mirroring when a directory is
> > replaced by a symlink
> > * Useless work done to mirror old snapshots that nobody needs, while
> > the latest snapshot has the highest business value
> > * Useless stat() calls slowing down the mirroring significantly
> > * They also introduced multi-threaded mirroring so that more than one
> > file is copied at a time and more than one subdirectory is scanned at
> > a time - this helps with directories containing many small files
> >
> > Due to the above, I absolutely cannot recommend using the unpatched
> > cephfs-mirror and would rather (if you don't want to contact croit for
> > patches) ask you to switch to rsync with some wrapper script that
> > runs, in parallel, multiple copies operating on different
> > subdirectories.
> >
> > On Fri, Mar 21, 2025 at 12:48 AM Vladimir Cvetkovic
> > <vladimir.cvetko...@hostpapa.com> wrote:
> >>
> >> Hi everyone,
> >>
> >> We have two remote clusters and are trying to set up snapshot mirroring.
> >> Our directories  are in a syncing state but never seem to actually sync
> the
> >> snapshots.
> >> Snapshot mirroring seems way to slow for large file systems - does
> anyone
> >> have a working setup?
> >>
> >> Thanks!
> >> _______________________________________________
> >> ceph-users mailing list -- ceph-users@ceph.io
> >> To unsubscribe send an email to ceph-users-le...@ceph.io
> >
> >
> >
>
>
> --
> Daniel Schreiber
> Facharbeitsgruppe Systemsoftware
> Universitaetsrechenzentrum
>
> Technische Universität Chemnitz
> Straße der Nationen 62 (Raum B303)
> 09111 Chemnitz
> Germany
>
> Tel:     +49 371 531 35444
>


-- 
Alexander Patrakov
_______________________________________________
ceph-users mailing list -- ceph-users@ceph.io
To unsubscribe send an email to ceph-users-le...@ceph.io

Reply via email to