Forum Discussion

mbrandon32's avatar
mbrandon32
Icon for Cirrus rankCirrus
May 20, 2024

GSLB - Monitoring LTM VIP load balancing via iRule

In one of our environments we are configuring a single LTM VIP and load balancing multiple applications via an iRule. We currently have other LTM environments integrated via iQuery with our GTM for GSLB configuration and monitoring.

Is there a way to monitor the VIP at the GTM level via iQuery that would give a true back-end pool status? Since, let's say, we are load balancing 100 different applications via a single VIP, if 99 of them went offline, the VIP would still show as ONLINE/GREEN.

Or would we even go as far as integrating via iQuery and adding a dependency monitor of the pool itself instead?

application delivery
gslb
gslb servers
GSLBPOOL
gtm
GTM DNS

6 Replies

Sort By
  • zamroni777's avatar
    zamroni777
    Icon for Nacreous rankNacreous
    May 20, 2024

    GTM serves DNSlayer that its dns response only contains ip addresses without any L4 port numbers.
    in case the GTM pool member is LTM VS, then GTM response contains their VIP address (without the L4 port number).

    as work around, you can make dummy LTM VS (using same VIP but different ports or different source address) containing monitors to each 100 applications, hence it will be marked Down if any of the monitor fails.
    then bind this new LTM VS to the GTM pool, replacing the real (non dummy) LTM VS.

  • Amine_Kadimi's avatar
    Amine_Kadimi
    Icon for MVP rankMVP
    May 21, 2024

    How is the LTM VIP green if you are using iRule to load balance, does it have a default pool which is green?

    • mbrandon32's avatar
      mbrandon32
      Icon for Cirrus rankCirrus
      May 21, 2024

      No default pool. We use an Address List for destination IP.

      • Amine_Kadimi's avatar
        Amine_Kadimi
        Icon for MVP rankMVP
        May 23, 2024

        So the VIP itself is blue and not green, the GTM will report it as is because using iquery the system fetches the LTM virtual server status from the LTM without further monitoring.

        You need to either implement some sort of monitoring on the LTM, but this is not be possible in your case unless you add a default pool (with monitoring enabled) to the VS, or to implement another type of monitoring except iquery on the GTM which will directly monitor the backend application.

  • F5_Design_Engineer's avatar
    F5_Design_Engineer
    Icon for MVP rankMVP
    May 21, 2024

    You should try configuring Prober Pools.

    GSLB prober pools monitor health probes of the sync group servers to gather health and performance data about the sync group server resources.

    All of the resources on your network are associated with a data center. BIG-IP DNS consolidates the paths and metrics data collected from the servers, virtual servers, and links in the data center. BIG-IP DNS uses that data to conduct load balancing and route client requests to the best-performing resource based on different factors. 

     DNS uses this data to optimize its load balancing decisions.

    You can view the GSLB prober pools configured for the DNS sync groups you manage by going to Configuration > DNS > GSLB > Prober Pools From there, click a pool to edit it, or click Create to set up a new pool.

    Benefits of F5 GSLB prober pools include:

    • Faster app routing
    • More efficient cloud scaling
    • Optimized cloud migration
    • Maximized app health, optimization, and protection 

    BIG-IP® DNS performs intelligent probing of your network resources to determine whether the resources are up or down. In some circumstances, for example, if your network contains firewalls, you might want to set up device-specific probing to specify which BIG-IP systems probe specific servers for health and performance data.

    Prober selection and fallback options are available to provide a variety of monitoring choices for data centers and servers. Allowing the customization of prober selection for a particular environment may be useful in reducing false negatives due to network configuration issues and outages. The prober selection option specifies which BIG-IP system(s) are preferred when selecting a local prober. The prober fallback option controls what the BIG-IP system does when not enough probers of the type specified are available.