The cluster safety job was initially intended to detect splits in a cluster. However, as it turns out, it also detects when a NON-CLUSTERED version of Confluence has been unintentionally deployed more than once. When these two or more non-clustered versions attempt to write to the same database simultaneously, the cluster safety job reports this error:
2007-03-19 19:43:30,020 FATAL [DefaultQuartzScheduler_Worker-0] [confluence.cluster.safety.ClusterPanicListener] handleEvent Fatal error in Confluence cluster: Database is being updated by an instance which is not part of the current cluster. You should check network connections between cluster nodes, especially multicast traffic.
The cluster specific part of this message should be ignored for users who are using non-clustered versions. However, users must address the underlying issue: that somehow two instances of Confluence have been deployed that are are both writing to the database elected for Confluence.
In future, we should rename the cluster safety job and make it throw the right johnson message depending on whether clustering has been enabled or not.