You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Apr 23, 2020. It is now read-only.
Recently I noticed on some exhibitor nodes, the backups directory, /var/exhibitor/backups, isn't deleting directories after the max save time (which is set to 1 hour). There are directories dating back 18 months, with the vast majority being empty (99.99% empty). This is causing some space issues on the nodes. I was planning on just deleting the empty directories, but wanted to reach out to see if anyone knew why the directories weren't being deleted.
I did look in the logs, but haven't seen anything to indicate a failure occurred while trying to delete them. Every log statement (in the current and rollover logs) is just a slight variation of this:
@400000005a1cffb02ac5c694 INFO com.netflix.exhibitor.core.activity.ActivityLog Cleanup task completed [pool-2-thread-161253]
The text was updated successfully, but these errors were encountered:
I was looking into this further and it seems like this is a bug. The backups are creating files following the path exhibitor backup dir / zookeeper data log file name / zookeeper data log file modified time. The FileSystemProvider however only deletes the file and not the parent directory.
I'm not sure if perhaps zookeeper used to maintain the same file and name so deleting the directory wasn't needed but this at least isn't true now.
Recently I noticed on some exhibitor nodes, the backups directory, /var/exhibitor/backups, isn't deleting directories after the max save time (which is set to 1 hour). There are directories dating back 18 months, with the vast majority being empty (99.99% empty). This is causing some space issues on the nodes. I was planning on just deleting the empty directories, but wanted to reach out to see if anyone knew why the directories weren't being deleted.
I did look in the logs, but haven't seen anything to indicate a failure occurred while trying to delete them. Every log statement (in the current and rollover logs) is just a slight variation of this:
The text was updated successfully, but these errors were encountered: