|
I have a 3 cluster multisite replication test environment (Ceph pacific 16.2.9). 3 zones in the zonegroup, each zone has a unique bucket and a bucket policy that replicates that bucketq to the other 2 zones. Each bucket has 1 or 2 small objects.
The radosgw in the master zone is constantly showing up at the top of the CPU usage, consuming anywhere from 50-100% of CPU, with no activity in any of the other 2 zones. Debug level is set to 0/1 because otherwise the logs end up getting flooded and filling
up the /var/log disk partition.
If I turn up logging to 1/5, it is a constant stream of requests for "GET /admin/log?type=metadata&id=XX...."
This seems like something is wrong, but sync status shows no problems on any of the zones.
Any suggestions or ideas what is going on?
|
_______________________________________________ Dev mailing list -- dev@xxxxxxx To unsubscribe send an email to dev-leave@xxxxxxx
