QUESTION: When I try to reach the link http://\[CSN IP]:8090/services/csn/backup/ [http://10.50.152.10:8090/services/csn/backup/], I get a 503. What's happening?
ANSWER: The problem is a known bug in CSN 3.x and 6.5, and it is fixed in CSN 7.0. What is happening is that any error message in /var/log/caringo/castor.log or cr.log is causing us to change a file (fifo.log) which is currently being stored in a location that is being monitored for CSN backups. ie, if that file changes, we think we need to generate a backup. That is causing numerous extra backups that are not necessary. The system, when you click on that link, is trying to load all of these backups and is timing out. The fix is to truncate the list of backups that are being loaded. The script will allow you to remove the extra backups as well, only keeping the most current backup. That is my suggestion.
Download the following to the CSN in the /root folder: https://support.cloud.caringo.com/tools/caringo-support-tools.tgz
unpack it:
cd;
tar zxvf caringo-support-tools.tgz
Then run:
/root/dist/csn-modify-saveset.sh -w
It will apply the patch with the -w option, and then ask you questions verifying that you want to truncate the backup file listing, and then even allow you to remove the old backups from the cluster itself (recommended if you don't need to restore a backup right now).