Details
-
Improvement
-
Status: Open
-
Minor
-
Resolution: Unresolved
-
4.7
-
None
-
None
-
None
Description
In cases of partial failure where a node is still connected to ZooKeeper but failing – e.g. bad disk, bad memory, etc. – it would be nice to have a quick UI link or command-line script to remove the node from clusterstate.json quickly.
We've had partial failures where we couldn't SSH into the box but the VM was still running and connected to ZooKeeper. In these cases, we've had to power the machine down from the ILO in order to get it out of clusterstate.json.
Having something handier in such outages would be great.
Attachments
Issue Links
- is related to
-
SOLR-5128 Improve SolrCloud cluster management capabilities
- Open