On 10/19/18 1:37 PM, Arvydas Opulskis wrote:
Yes, that's understandable, but question was about "transit period" when at some point we had part of RGW's upgraded and some of them were still in Jewel. At that time we had a lot of complains from S3 users, who couldn't access their buckets randomly. We did several upgrades in last years and it was first time when old RGW's were not able to work with buckets "touched" with new version RGW's.
You should plan downtime and notify your clients about it. DreamHost (DreamObjects service) did it at the beginning of the year.
Or you should fast upgrade/restart - this is 10-30 seconds for LB scheme. k _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
