Forum Discussion

Matthew_59082's avatar
Matthew_59082
Icon for Nimbostratus rankNimbostratus
Jul 05, 2011

Both LTM HA Peers going active.

Hello,

 

 

I am having a problem with two HA peers, both are LTMs running 10.1.0. Both peers will go active and stay active no matter what.

 

 

I have the HB network on a dedicated link between the two LTMs, they can ping and config sync accross the HB network, this works everytime. I have to keep one LTM forced offline, and if I release it, it will go standby then active. If I force one to standby a few seconds later it goes back to active.

 

 

Virtual servers continue to work even when they both go active, I think this is because the second unit is not second a gratuitous arp. One unit has LVM, the other is on partitions thats the only difference between the LTMs, otherwise my setup is mirroring the setup used in another DC and it works fine.

 

 

How can I get a better view of the HA process? The LTM log simply gives a line "Standby" then "Active" but not the reason why.

 

 

 

Thanks.
management
monitoring
  • nathe's avatar
    nathe
    Icon for Cirrocumulus rankCirrocumulus
    Jul 05, 2011
    Matthew

     

     

    Have you tried running a tcpdump on both LTMs to make sure the network heartbeat is reaching the other device?

     

     

    Also, is Redundancy State Preference set on both LTMs? Could you have both set as Active?

     

     

    Hope this helps

     

    N
  • hooleylist's avatar
    hooleylist
    Icon for Cirrostratus rankCirrostratus
    Jul 05, 2011
    If the configuration is okay and the network between the two units is reachable, then it might be a resource issue where the failover state daemon isn't able to send heartbeats to the peer. You could open a case with F5 Support to get help diagnosing the issue.

     

     

    Aaron
  • Troy_93881's avatar
    Troy_93881
    Icon for Nimbostratus rankNimbostratus
    Jul 05, 2011
    Mathew,

     

     

    Is it possible for you to use or try the "hardwire" failover?
  • Matthew_59082's avatar
    Matthew_59082
    Icon for Nimbostratus rankNimbostratus
    Jul 11, 2011
    Thanks for all the replies...

     

     

    I haven't used tcpdump yet, that's a good idea I'll do that now. I'm looking for port tcp port 1026?

     

     

    The units are idle, so i don't think they're out of resources, unfortunately I cannot use hardwire failover either.
  • Matthew_59082's avatar
    Matthew_59082
    Icon for Nimbostratus rankNimbostratus
    Jul 11, 2011
    I can see the HA messages going back and forth.

     

     

    16:24:48.793179 IP 192.168.0.1.43364 > 192.168.0.2.cap:

     

    16:24:48.810475 IP 192.168.0.2.34891 > 192.168.0.1.cap:

     

    16:24:48.893170 IP 192.168.0.1.43364 > 192.168.0.2.cap:

     

    16:24:48.910457 IP 192.168.0.2.34891 > 192.168.0.1.cap:

     

     

    The LTMs just seem to be ignoring each other.
  • youssef1's avatar
    youssef1
    Icon for Cumulonimbus rankCumulonimbus
    May 04, 2018

    The start URL have to be the url that host Login Page.

     

    In fact is URL that will be trigged your SSO. when you will be in the login page, just retrieve the url where you are (Via F12 to be sure or you can see it in your web bar)...

     

    Regards