Discussion:
[ceph-users] Minimal downtime when changing Erasure Code plugin on Ceph RGW
Charles Alva
2018-12-07 04:45:39 UTC
Permalink
Hi All,

I would like to perform a proof of concept on changing a running Ceph
cluster which is using non-default erasure code plugin back to jerasure
plugin on its default.rgw.buckets.data pool. I could not find a
documentation on how to achieve this with minimal downtime. I know changing
erasure code plugin and profile would require us to migrate the data inside
the old pool to the new pool.

Is there a better way to change the erasure code plugin with minimal
downtime as possible for Ceph RGW with hundred millions of objects?

Since `rados cppool` is deprecated, I'm considering in using cache tier
method as explained at https://ceph.com/geen-categorie/ceph-pool-migration/.
Is cache tier method feasible?


Kind regards,

Charles Alva
Sent from Gmail Mobile

Loading...