Uploaded image for project: 'Confluence Data Center'
  1. Confluence Data Center
  2. CONFSERVER-9782

Document maintaining an offsite backup instance for failover

    • We collect Confluence feedback from various sources, and we evaluate what we've collected when planning our product roadmap. To understand how this piece of feedback will be reviewed, see our Implementation of New Features Policy.

      NOTE: This suggestion is for Confluence Server. Using Confluence Cloud? See the corresponding suggestion.

      Some organisations require the option to switch all mission-critical applications to an offsite backup in the event that the primary instance goes down. Please document the basics of setting this up:

      1. Install Confluence to a backup server using a development key that is not running
      2. Setup a load balancer front end that can test if Confluence is responsive on the primary and backup servers
      3. Setup a database backup system that transfers the database contents to the backup server database, eg nightly or hourly
      4. Setup a script so that the load balancer will start the backup instance, stop the periodic database transfers and redirect requests in the event that the primary goes down

      If there's a better method then we should document that instead.

            [CONFSERVER-9782] Document maintaining an offsite backup instance for failover

            Great Suggestion,

            1) HA is needed in many organizations for production purposes.
            I'm not talking only about Load-balancing but a real HA.
            The need for system which would shutdown unresponsive application (What would be your implementation for monitoring?)
            and start another instance on another server automatically.

            2) The perfect solution would be to run 2 instances simultaneously for load balancing and HA (so we could use hardware Load balancer, for example).
            But my understanding that your current caching technique won't allow this.

            Leon Kolchinsky added a comment - Great Suggestion, 1) HA is needed in many organizations for production purposes. I'm not talking only about Load-balancing but a real HA. The need for system which would shutdown unresponsive application (What would be your implementation for monitoring?) and start another instance on another server automatically. 2) The perfect solution would be to run 2 instances simultaneously for load balancing and HA (so we could use hardware Load balancer, for example). But my understanding that your current caching technique won't allow this.

            Matt Ryall added a comment -

            Dave's suggestion looks good. One simple way to provide recovery is by reconfiguring DNS in the event of a problem (and ensuring DNS TTL setting is low, so clients update from the server frequently).

            In addition, this page should link to the documentation for:

            • how to back up Confluence
            • how to install Confluence
            • where important files are kept and what they do.

            We should also note that combining this information into a comprehensive disaster recovery plan is the responsibility of any organisation that has these requirements. The exact methods used and the requirements for availability vary greatly between organisations, so our document provides only a loose outline of what is possible, not any kind of recommendation or support agreement.

            Matt Ryall added a comment - Dave's suggestion looks good. One simple way to provide recovery is by reconfiguring DNS in the event of a problem (and ensuring DNS TTL setting is low, so clients update from the server frequently). In addition, this page should link to the documentation for: how to back up Confluence how to install Confluence where important files are kept and what they do. We should also note that combining this information into a comprehensive disaster recovery plan is the responsibility of any organisation that has these requirements. The exact methods used and the requirements for availability vary greatly between organisations, so our document provides only a loose outline of what is possible, not any kind of recommendation or support agreement.

              Unassigned Unassigned
              david.soul@atlassian.com David Soul [Atlassian]
              Votes:
              3 Vote for this issue
              Watchers:
              3 Start watching this issue

                Created:
                Updated:
                Resolved: