Forum Discussion

Anthony's avatar
Anthony
Icon for Nimbostratus rankNimbostratus
Dec 10, 2018

Disconnected, Red vs Blue - whats the difference?

I'm having real trouble reconnecting an HA pair of VE devices. I've run through various troubleshooting guides, reset device trusts, recreated groups, pinged all addresses, confirmed NTP sync, port lockdown etc, but I get 1 device saying Disconnected in RED, and the other Disconnected in BLUE - what is the difference?

 

When I go to Device Management > Overview, the device with the RED disconnected shows the other device in an Unknown State. Which the "BLUE" device reports the state quite happily.

 

So I'm guessing there is a communication issue between the devices in the direction where the RED is showing, but would like some ideas of where this is likely to be as all IPs seem ok and can communicate.

 

Any help would be greatly appreciated!

 

Regards, Anthony

 

3 Replies

  • Hey Anthony

    Can you perform a configuration sync?

    Have you checked so that the BIG-IP devices are listening over the correct failover unicast address/port? Run the following command on both devices:

    netstat -na | grep 1026

    The output should look like this:

    [root@bigip02:Active:Changes Pending] log  netstat -na | grep 1026
    
    udp        0      0 192.168.1.32:1026           0.0.0.0:*
    udp        0      0 10.10.10.32:1026            0.0.0.0:*
    

    This setting is configured per device under: Device Management ›› Devices ›› [BIG-IP Hostname] ›› Failover Network.

    I think that if it's missing it could present the scenario you have.

    I hope this helps! 🙂

  • I've just disabled the HA interface which sent the Disconnected state to RED, which seems to mean that it cannot communicate with its peer. When I turned this back on, it eventually went BLUE, so it could communicate but it couldn't sync. That seems to be the difference with the colour. Issue still exists with the port 6699 being unavailable on one of the devices. As these are VE's we're going to try moving the one with the problem to a different host to force and ARP which will hopefully clear any issues with ports being held open.