Hi Isaiah,
A simple solution for multi-site redundancy is to have two nearby sites
with < 3ms latency and setup crush map [0] for datacenter level
redundancy instead of the default host level.
Performance was adequate in my testing for large number of small files
if the latency between all nodes were kept below 3 ms. Of course it also
depends on your application.
Ceph fs snapshot mirroring is asynchronous, so your application would
need to handle the logic of switching to the replica node, operating
with some missing data in a degraded state, synchronizing changes back
to primary after it comes online and switching back. Too complicated IMHO.
[0]: https://docs.ceph.com/en/quincy/rados/operations/crush-map/
Kind regards,
Pavin Joseph.
On 28-Dec-22 11:27 AM, Isaiah Tang Yue Shun wrote:
Hi all,
From the documentation, I can only find Ceph Object Gateway multi-site
implementation. I wonder is it if we are using CephFS, how can we achieve
active-active setup for production?
Any input is appreciated.
Thanks.
Regards,
Isaiah Tang
_______________________________________________
ceph-users mailing list -- ceph-users@ceph.io
To unsubscribe send an email to ceph-users-le...@ceph.io
_______________________________________________
ceph-users mailing list -- ceph-users@ceph.io
To unsubscribe send an email to ceph-users-le...@ceph.io
