Description
I found this while testing the changes in SOLR-12739. The ZkController creates a SolrCloudManager with a new CloudSolrClient instance. However, no connection or read timeout is set for this CloudSolrClient. Therefore, when the policy framework tries to read freedisk metrics for a dead node (this is another bug), it just hangs forever causing collection APIs to timeout.
Attachments
Issue Links
- blocks
-
SOLR-12739 Make autoscaling policy based replica placement the default strategy for placing replicas
- Closed