Uploaded image for project: 'Jira Data Center'
  1. Jira Data Center
  2. JRASERVER-61196

Please make our database failover details clearer for JIRA Data Center

XMLWordPrintable

    • 2
    • 4
    • We collect Jira 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 JIRA Server. Using JIRA Cloud? See the corresponding suggestion.

      Currently we don't list anything about database failover in JIRA, or Data Center - we do have https://confluence.atlassian.com/display/enterprise/disaster+recovery+guide+for+jira which discusses cold standby but I wasn't able to easily tie that back to database failovers.

      Essentially if the database underneath JIRA changes it's going to crash - this request is to document that somewhere and also that if setting up a DB failover it has to be done in JDC.

            [JRASERVER-61196] Please make our database failover details clearer for JIRA Data Center

            Matt Doar added a comment -

            My experience is that if the cname of the database server is changed to a different database host, it took 15 minutes for our Jira DC 4 node cluster to recover.

            Matt Doar added a comment - My experience is that if the cname of the database server is changed to a different database host, it took 15 minutes for our Jira DC 4 node cluster to recover.

            Damian Wheeler (Otago) added a comment -

            That might depend on your database of choice and your HA configuration of the database. This arena can be quite complex depending on what level of resilience the site is aiming for. It probably isn't something that Atlassian would want to seem responsible for if they start documenting that. All that tomcat should care about is a jdbc connection pool connecting to an available data source. If that is Oracle RAC or just a Dataguard primary and standby, or Active dataguard with automated failover, or PostgreSQL with a proxy layer handing failovers, or SQL Server with Always ON, etc etc. Too many variables for Atlassian to describe and if you want to run a complex cluster on-premise, then you should have the database expertise to go with it. Do you want to talk about multi-site replication?
            Database technologies describe how to do HA properly, all JIRA needs is the correct JDBC connection string which I think is already mentioned in the documentation. My experience is that it copes with a little database downtime and recovers just fine with it is available again. It doesn't "crash".

            Damian Wheeler (Otago) added a comment - That might depend on your database of choice and your HA configuration of the database. This arena can be quite complex depending on what level of resilience the site is aiming for. It probably isn't something that Atlassian would want to seem responsible for if they start documenting that. All that tomcat should care about is a jdbc connection pool connecting to an available data source. If that is Oracle RAC or just a Dataguard primary and standby, or Active dataguard with automated failover, or PostgreSQL with a proxy layer handing failovers, or SQL Server with Always ON, etc etc. Too many variables for Atlassian to describe and if you want to run a complex cluster on-premise, then you should have the database expertise to go with it. Do you want to talk about multi-site replication? Database technologies describe how to do HA properly, all JIRA needs is the correct JDBC connection string which I think is already mentioned in the documentation. My experience is that it copes with a little database downtime and recovers just fine with it is available again. It doesn't "crash".

              wthompson Wazza
              dcurrie@atlassian.com Dave C
              Votes:
              8 Vote for this issue
              Watchers:
              13 Start watching this issue

                Created:
                Updated: