F5 is upgrading its customer support chat feature on My.F5.com. Chat support will be unavailable from 6am-10am PST on 1/20/26. Refer to K000159584 for details.

Forum Discussion

Jay_Wong_141181's avatar
Jay_Wong_141181
Icon for Nimbostratus rankNimbostratus
Jan 12, 2014

how to get what node is stuck with an F5 virtual - Persist type: dest_addr

Hi,

 

There are a few F5 virtuals used in my firm. Let's say one is called "virtual_persist". Behind that, we have two nodes - A & B. The problem is that the virtual directs request to node A exclusively for a while and then fails over to node B. It will impact our application when fail-over happens without us knowing it. I'm writing a monitor script aiming to get what the current destination address is. But I have no idea. Can you shed some light? Any advice will be appreciated.

 

3 Replies

  • BinaryCanary_19's avatar
    BinaryCanary_19
    Historic F5 Account

    If you want traffic to only go to Node B when Node A is actually down, use Priority Groups.

     

    http://support.f5.com/kb/en-us/products/lc_9_x/manuals/product/lc_config_10_2/lc_pools.html?sr=34369937

     

    Look for the version of the manual that applies to your software version.

     

    You should also be able to configure custom SNMP alerts to send you an email/or raise an SNMP trap whenever Node A or Node B goes down.

     

    Search http://ask.f5.com for Custom SNMP Alerts

     

  • It's not decided by me. The F5 services are provided by another team so I have limited control of it. What I want actually is a command or api that can query what node the virtual is pointing to currently. In my mind, it looks like this:

     

    f5query -virtual "virtual_persist"

     

    Checking... Two node totally A & B Currently virtual_persist is stuck with A

     

    So are you saying we don't have such thing?

     

  • BinaryCanary_19's avatar
    BinaryCanary_19
    Historic F5 Account

    The concept of "which node the virtual is pointing to" is not accurate.

    Virtuals do not point somewhere. They load balance connections to configured nodes (pool members) according to the algorithm you have chosen (default is round robin).

    If you are using Destination address persistence, what it means is that "destination Address" of incoming connections is used as a key into the lookup table that stores persistence information.

    So, connection (or packet) going to 10.1.2.3 comes in, F5 does a lookup in persist table for "10.1.2.3", if it finds an entry, it directs the connection or packet to whatever the persist table says it should send it to. If there is no entry for that key, it will chose a new destination according to the algorithm configured (again, default is round robin) and then add an entry to the persist table with this key, then send the connection/packet to the chosen destination.

    To examine this persistence table, you can do soemthing like:

    tmsh show ltm persistence persist-records

    The command has a rich set of options that allows you to filter the results by source or destination address etc. You can find more details in the "TMSH Reference" for your software version at http://ask.f5.com