We have tried to change one of the zookeeper node out of 6 to observer. After executing the re-config command the observer node is not reachable. It is only occurring in production cluster and not occurring in a test cluster with same configuration. Have anyone faced this issue?

I am currently dealing with an issue where occasionally a system using ZooKeeper will generate are large number of snapshots and transaction log files. These take up disk space that we would like to free up as quickly as possible.

I have a znode as this:

zk version: 3.8.3 first of all my zk cluster crushed, and i want to findout the reason. i dump the snapshot and datalog to my test server but it can’t startup either then i delete the datalog file and restart , it succeed BUT when i stoped the instance and restart , it failed, can’t […]

I want to compile c client for zookeeper, I cloned the codes and checkout it to tag 3.8.4. Then I did as follows:
1 install libcppunit-dev

I use ZooKeeper to acquire a distributed lock. If I obtain the lock but the task takes too long to execute, will the lock be released prematurely?

I am running ZooKeeper latest stable 3.8.4 release. I installed it in /usr/local/zookeeper/ and I configured /usr/local/zookeeper/conf/zoo.cfg to specify