Created on 03-09-2016 07:54 AM - edited 09-16-2022 03:08 AM
Created 03-09-2016 06:33 PM
The autopurge.snapRetainCount configuration setting is described here.
https://zookeeper.apache.org/doc/r3.4.8/zookeeperAdmin.html#sc_advancedConfiguration
Old snapshots/logs represent earlier points in time for the state saved in a ZooKeeper cluster, similar to a backup. Tuning the autopurge.snapRetainCount represents a trade-off between disk space consumption and flexibility to restore back to earlier points in time.
In practice, I have not needed to restore state to an earlier point in time like this. Typical applications of ZooKeeper use it for transient coordination data that can be recreated from first principles easily in the event of a disaster, so going back to an earlier point in time generally isn't very important. For that reason, I haven't needed to tune up the retention count.
Created 03-09-2016 06:33 PM
The autopurge.snapRetainCount configuration setting is described here.
https://zookeeper.apache.org/doc/r3.4.8/zookeeperAdmin.html#sc_advancedConfiguration
Old snapshots/logs represent earlier points in time for the state saved in a ZooKeeper cluster, similar to a backup. Tuning the autopurge.snapRetainCount represents a trade-off between disk space consumption and flexibility to restore back to earlier points in time.
In practice, I have not needed to restore state to an earlier point in time like this. Typical applications of ZooKeeper use it for transient coordination data that can be recreated from first principles easily in the event of a disaster, so going back to an earlier point in time generally isn't very important. For that reason, I haven't needed to tune up the retention count.
Created 03-10-2016 05:47 AM
Thanks for your reply @Chris Nauroth