After an upgrade of our small cluster (two indexers, one search head, RF=SF=2) to 6.6.0, 5 indexes are marked as not searchable.
There are 0 "fix up" type tasks. I have tried gracefully restarting both indexers again (one at a time) and no luck there. Any help is appreciated. Thanks.
no Fix-up tasks
OK the situation appears OK now. This may not be a commonly happening case. I ran the commands below to find the buckets status ;
i) REST call https://ClusterMaster:8089/services/cluster/master/buckets?filter=replication_count%3C2&filter=froze...
List of buckets with "frozen=false" & "replication_count < 2" (RF=SF=2 deployment)
ii) After I get the list of buckets from above, checked individual buckets using https://ClusterMaster:8089/services/cluster/master/buckets/{BID OF BUCKET IN ABOVE LIST}
Strangely there were some buckets without PEER information, which appears that the cluster master has some buckets information that its peers do not have.
Then ran remove_all which worked around the issue :
curl -k -u admin:password https://:8089/services/cluster/master/buckets//remove_all -X POST
OK the situation appears OK now. This may not be a commonly happening case. I ran the commands below to find the buckets status ;
i) REST call https://ClusterMaster:8089/services/cluster/master/buckets?filter=replication_count%3C2&filter=froze...
List of buckets with "frozen=false" & "replication_count < 2" (RF=SF=2 deployment)
ii) After I get the list of buckets from above, checked individual buckets using https://ClusterMaster:8089/services/cluster/master/buckets/{BID OF BUCKET IN ABOVE LIST}
Strangely there were some buckets without PEER information, which appears that the cluster master has some buckets information that its peers do not have.
Then ran remove_all which worked around the issue :
curl -k -u admin:password https://:8089/services/cluster/master/buckets//remove_all -X POST