Deployment Architecture

With no fixup tasks pending, why are the Clustering UI states replication and search factors are not met?

Explorer

When checking on my replication and search factors on my cluster master, it states they are not met. There are a number of different indexes containing various combinations of buckets which do not meet either the replication factor, or search factor, or both. When I click on the "Bucket Status" button on the "Indexes" tab, there are no pending fixup tasks. How do I get my cluster to meet its replication and search factors again?

1 Solution

Splunk Employee
Splunk Employee

The situation you described occurs only rarely. To correct the situation you may do the following.

  • From a browser, navigate to the following location on your cluster master to get a list of the non-frozen buckets which are not meeting the replication factors:

https://servername:8089/services/cluster/master/buckets?filter=replication_count%3C3&filter=frozen=f...

https://servername:8089/services/cluster/master/buckets/_audit~120~67601672-E174-460E-BDDE-FDF32CDCB...

  • If you see force_roll set to 1, and the bucket is not meeting replication factor, the bucket needs to be fixed manually.
  • Otherwise you may put use the following curl call to put the identified bucket into the fixup queue:

curl -k -u admin:admin https://servername:8089/services/cluster/master/buckets/_audit~120~67601672-E174-460E-BDDE-FDF32CDCB... -X POST

View solution in original post

Splunk Employee
Splunk Employee

You can also try restarting the cluster-master. This should clear up any stale fix-up tasks.

0 Karma

Splunk Employee
Splunk Employee

The situation you described occurs only rarely. To correct the situation you may do the following.

  • From a browser, navigate to the following location on your cluster master to get a list of the non-frozen buckets which are not meeting the replication factors:

https://servername:8089/services/cluster/master/buckets?filter=replication_count%3C3&filter=frozen=f...

https://servername:8089/services/cluster/master/buckets/_audit~120~67601672-E174-460E-BDDE-FDF32CDCB...

  • If you see force_roll set to 1, and the bucket is not meeting replication factor, the bucket needs to be fixed manually.
  • Otherwise you may put use the following curl call to put the identified bucket into the fixup queue:

curl -k -u admin:admin https://servername:8089/services/cluster/master/buckets/_audit~120~67601672-E174-460E-BDDE-FDF32CDCB... -X POST

View solution in original post