ConfigHandler dead locks in K8 when delete pod is used
Description
I believe this is related to the connection pool implementation as the behavior is seen in both the ChannelConnectionPool and the DataFlowConnectionPool; the SingleConnectionPool does not seem to have the same issue.
Note this only appears to happen in K8 environments.
Environment
None
Gerrit Reviews
None
Release Notes Description
None
Activity
Show:
Jeffry Morris May 4, 2021 at 11:49 PM
ClusterNode.GetClusterMap() lacked cancellation support and thus never timed out.
Fixed
Pinned fields
Click on the next to a field label to start pinning.
I believe this is related to the connection pool implementation as the behavior is seen in both the ChannelConnectionPool and the DataFlowConnectionPool; the SingleConnectionPool does not seem to have the same issue.
Note this only appears to happen in K8 environments.