pravega / zookeeper-operator

Kubernetes Operator for Zookeeper
Apache License 2.0
366 stars 203 forks source link

Adding a node may fail and there is no retry #552

Open runkecheng opened 1 year ago

runkecheng commented 1 year ago

Description

java -Dlog4j.configuration=file:"$LOG4J_CONF" -jar /opt/libs/zu.jar add $ZKURL $MYID $ZKCONFIG $DYNCONFIG

This operation may fail and there will be no retry, resulting in the node never joining the cluster again. The phenomenon is usually: when creating a three-node cluster, the second node is not ready, use zkCli.sh to check the configuration where the entry of the second node is missing.

Importance

Location

Suggestions for improvement

Add a check: after adding a node, "zu.jar get" can be used to check if the current node is included in the peer list