Since upgrading SRX 340 from 15.1XD60 we are seeing a strange behaviour where the Fab 0 link does down taking down the cluster , rebooting the secondary node temporarility restores the cluster heath with heartbeat exchanged but soon the secondary node goes back into "ineligible" state and eventually into disabled mode. The device was RMA'ed once with the same behaviour shown on 15.1X49-D75 version. The SRX are connected back to back the cables were replaced. Tried to disable the fabric link monitoring but that didnt help either. https://kb.juniper.net/InfoCenter/index?page=content&id=KB22717 this is exactly the same behaviour we are experiencing. Upgrading the SRX to a newer recommended version temporary solves the issue. JTAC is recommending to perform a clean wipe of the secondary node and having it rejoin the cluster with both nodes rebooted, this was exactly the same steps peformed twice before but the issue has resurfaced. This only happens on the seconday node, has any one experienced this behavior?
↧