On Thu, Jan 23, 2014 at 4:27 PM, Greg Poirier <greg.poirier@xxxxxxxxxx> wrote: > Hello! > > I have a great deal of interest in the ability to version objects in buckets > via the S3 API. Where is this on the roadmap for Ceph? This still needs to be prioritized. There was a discussion about it in the latest CDS, but there wasn't any follow up. We're certainly interested in having this feature, however, we still need to figure out how the implementation is going to look like. > > This is a pretty useful feature during failover scenarios between zones in a > region. For instance, take the example where you have a region with two > zones: > > us-east-1 > us-west-1 > > In the us region, east is master, west is secondary. > > In the cast where us-east-1 fails, we want to failover to us-west-1. Let us > assume that we have versioned all of our objects in a bucket. This means > that every version of an object should map to a different object in Ceph. > So, in the case of a us-east-1 failure, clients could still write new > versions of an object to us-west-1 until us-east-1 becomes available again. > > I am working under the assumption that after the restoration of service of > us-east-1, objects written to us-west-1 will replicate to us-east-1, > however. Is that the case? Such a scenario should be feasible with the current set of tools that we developed for disaster recovery / geo-replication. > > I realize that this is just a convenience and that we could embed a version > number in our object names somehow, but that is somewhat less... clean? Plus > we cannot re-use code that was written against S3. :( > For each object that rgw stores it keeps a version tag. However this version is not ascending, it's just used for identifying whether an object has changed. I'm not completely sure what is the problem that you're trying to solve though. Yehuda _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
 |