User story #3966
open
Document a failover procedure for Rudder
Added by Dennis Cabooter about 11 years ago.
Updated 2 months ago.
Description
Is is possible to have a master and a slave Rudder server? Example. Our wiki's are physically located on a different location. All changes in our main wiki are synced nightly. In that case we have always a backup on a remote location. Would it be already possible or would it be an idea to be able to do this with the Rudder server as well?
- Category set to 13
- Status changed from New to Discussion
- Assignee set to Dennis Cabooter
- Priority changed from N/A to 2
Hi Dennis,
If I am correct, you would like to have a failover mechanism on the Rudder server, to have a backup copy of your main Rudder server in case of a failure ?
- Project changed from Rudder to 30
- Subject changed from Is is possible to have a master and a slave Rudder server? to Document a failover procedure for Rudder
- Category deleted (
13)
That would require having at least four elements:
- A master / slave or master / master ldap service (Quite easy with slapd)
- A master / slave PostgreSQL service (harder, especially because PostgreSQL takes quite a big load of hits because of the reporting)
- A new instance of the Rudder webapp.
- A synchronization of Rudder's variable filesystem part (/var/rudder/configuration-repository)
This would need to be documented. Does that seem reasonable to you Dennis ?
- Project changed from 30 to Rudder
- Category set to Documentation
- Assignee deleted (
Dennis Cabooter)
- Priority changed from 2 to N/A
Also available in: Atom
PDF
Updated by Matthieu CERDA 
Updated by 
Updated by 
Updated by 
Updated by