Hi Simon,
What about the idea of creating the cluster over two data centers?
Would it be possible to modify the crush map, so one pool gets
replicated over those two data centers and if one fails, the other
one would still be functional?
A stretched cluster is a valid approach, but you have to consider
several things like MON quorum (you'll need a third MON independent of
the two DCs) and failure domains and resiliency. The crush map and
rules can be easily adjusted to reflect two DCs.
Additionally, would it be possible to prioritize one data center
over the other?
This would allow saving data from site1 to a pool on site2 in case
of a disaster on site1, site2 would still have those Backups.
I don't think you can prioritize, maybe the primary-affinity of the
OSDs in the second DC can be reduced to control that a little. But in
a stretched cluster you don't have two pools that are replicated, the
pools are stretched over those two DCs.
Depending on the actual requirements you could for example have your
(replicated) cephfs_data pool with size 4 and use a ruleset that
stores 2 replicas in each DC. This also works with EC, of course. As
long as your MONs have quorum one DC can fail and the data would still
be available from the other DC.
Regards,
Eugen
Zitat von Simon Sutter <ssutter@xxxxxxxxxxx>:
Thanks Stefan,
First of all, for a bit more context, we use this ceph cluster just
for hot backups, so 99% write 1% read, no need for low latency.
Ok so the snapshot function would mean, we would have like a colder backup.
Just like a snapshot of a VM, without any incremental functionality,
which also means scheduled but huge transfer rates.
What about the idea of creating the cluster over two data centers?
Would it be possible to modify the crush map, so one pool gets
replicated over those two data centers and if one fails, the other
one would still be functional?
Additionally, would it be possible to prioritize one data center
over the other?
This would allow saving data from site1 to a pool on site2 in case
of a disaster on site1, site2 would still have those Backups.
We have a 10G connection with around 0.5ms latency.
Thanks in advance,
Simon
________________________________
Von: Stefan Kooman <stefan@xxxxxx>
Gesendet: Dienstag, 8. September 2020 11:38:29
An: Simon Sutter; ceph-users@xxxxxxx
Betreff: Re: Syncing cephfs from Ceph to Ceph
On 2020-09-08 11:22, Simon Sutter wrote:
Hello,
Is it possible to somehow sync a ceph from one site to a ceph form
another site?
I'm just using the cephfs feature and no block devices.
Being able to sync cephfs pools between two sites would be great
for a hot backup, in case one site fails.
It's a work in progress [1]. This might do what you want to right now:
[2]. Note: I haven't used [2] myself.
Gr. Stefan
[1]: https://docs.ceph.com/docs/master/dev/cephfs-mirroring/
[2]: https://github.com/oliveiradan/cephfs-sync
_______________________________________________
ceph-users mailing list -- ceph-users@xxxxxxx
To unsubscribe send an email to ceph-users-leave@xxxxxxx
_______________________________________________
ceph-users mailing list -- ceph-users@xxxxxxx
To unsubscribe send an email to ceph-users-leave@xxxxxxx
