Created 8/13/2014 aaron.enfield · Updated 5/17/2016 jamshid.afshar
- If you have a CSN in a development environment and want to reset the environment without re-installing the server's operating system, you can follow these steps:
ON THE SWARM NODES: 1. Format the Swarm nodes one by one using the terminal console utility menu on each node. Wait to reboot until you have made all of the changes on the CSN. It's important that the newly formatted volumes remain down until all nodes have been reformatted, otherwise old streams might migrate into the newly formatted cluster.
ON THE CSN- CSN'S AFTER CSN 7.X WILL NO LONGER INCLUDE THE CONTENT ROUTER SO YOU CAN IGNORE THE CR-RELATED STEPS BELOW. 2. Stop services: service cr-publisher stop; service cr-replicator stop;service scspproxy stop; service rsyslog stop 3. Remove the CR datastores on the CSN so the CR does not have old events: rm -rf /u01/var/opt/caringo/contentrouter/ 4. Remove all of the relevant logs (at a minimum, those below): rm -rf castor.log cr.log scspproxy.log csn.log 5. Remove CSN backups and references to them: rm /var/opt/caringo/csn/backup/manifest/.bootstrap /var/opt/caringo/csn/backup/manifest/savesets.errors /var/opt/caringo/csn/backup/manifest/savesets.default 6. Reboot the CSN 7. Once the CSN is back up, reboot the storage nodes
The recommendation if you are going from a development environment into production is to do a complete reinstallation of the CSN server's RHEL/ CentOS OS and CSN code, and then to also format all of the Swarm nodes' disks via the utility menu.