Forum Discussion

Mark_Malyj_2467's avatar
Mark_Malyj_2467
Icon for Nimbostratus rankNimbostratus
Jan 05, 2017

Redundant pair by adding a second BIG-IP VE to an existing system

I have a single BIG-IP v12.0.0 Virtual Edition VM installed in a vSphere 5.5 environment. I want to create a redundant pair by adding a second BIG-IP VE.

 

? How do I get the second VE to initially boot into standby mode?

 

I am following the article . It says I should “Connect the new BIG-IP Controller with a fail-over cable to the existing BIG-IP Controller and turn on the power.” That is for physical BIG-IP, and does not apply to VE. It also says “If the systems are going to be installed in physically separate locations, do not connect the new system to your network infrastructure until you ensure it will remain in standby mode after rebooting. For more information, refer to .”

 

That article refers to outdated “bigpipe” commands. So I decided to substitute equivalent tmsh commands from the article Bigpipe Mappings. I proceeded by first disabling all the NICs inside vSphere for my second VE, then I did an initial power up, and logged in as root from the vSphere console. I then ran these tmsh commands to try to force the second VE to boot into standby mode:

 

tmsh run /sys failover standby

 

tmsh save sys config partitions all

 

But how do I confirm that this VE is really in standby mode? I don't think it is, because this command:

 

tmsh run cm watch-sys-device

 

… it shows “failover_state: HA_ACTIVE” instead of “HA_STANDBY”. I also tried:

 

tail -f /var/log/ltm

 

… shows “Command: go standby tmsh”, but I don’t see a line that says just “Standby” as expected from How do you check if my bigip was forced offline or entered into failover.

 

? So, how do I get the second VE to initially boot into standby mode? After I do that, then I will re-enable the NICs inside vSphere, and I will continue with the redundant pair procedure.

 

~ Mark

 

application delivery
BIG-IP
config
design
devops
LTM

2 Replies

Sort By
  • Soda_Cup_148395's avatar
    Soda_Cup_148395
    Icon for Nimbostratus rankNimbostratus
    Jan 05, 2017

    Sounds like the bigger answer to your need is scheduling a maintenance window. Besides that, clustering works exactly the same as a physical f5, the article you reference is talking about the failover cable, which isn't used anymore.

     

    https://support.f5.com/kb/en-us/products/big-ip_ltm/manuals/product/tmos-implementations-12-0-0/3.html

     

    For the vm, I would disable the NICs and then run a config-sync and verify everything is okay. This can be done when it is forced offline if you like.

     

  • Mark_Malyj_2467's avatar
    Mark_Malyj_2467
    Icon for Nimbostratus rankNimbostratus
    Jan 05, 2017

    @Soda Cup, I already have one functioning Big-ip that I set up some time ago. I am fairly new to F5 configuration. I am uneasy about how to juggle the IP addresses that are already configured for the first device in with the new IPs for the second device, and how to create shared IPs. talks about that, but you say its not up to date.

     

    Your article only talks about creating a redundant pair starting from scratch, not about adding a second device to one that is already working well. Is there a procedure somewhere that spells that out?