Forum Discussion
hunghsingchin
Altocumulus
Altocumulusgslb prober pool round robin mechanism
I have a query about the GSLB prober pool. Suppose a virtual server (VS) references multiple health check monitors with different execution intervals, and this VS is linked to a prober pool that includes several GTM members from different data centers, utilizing a round-robin load balancing algorithm. I've just learned that the prober pool's round-robin mechanism is independent of the health check monitors' execution intervals, meaning that monitors can be driven by their own intervals to execute. What then drives the round-robin selection of prober pool members? My question might be a bit unclear, so let me give a specific example: Suppose a VS has three monitors with intervals of 3, 4, and 5 seconds, and at the 0-second mark, gtm1 is selected and executes all three monitors. Which prober pool member will execute the monitors at the 3rd, 4th, and 5th seconds, respectively? Is the prober pool's round-robin mechanism independent, or what drives it? Is it time-driven, and is this interval configurable by the user? and if prober pool member in different dcs，how does they know who will be choose this round and this time ?
- whisperer
MVP
Here is the only thing I was able to find out:
https://my.f5.com/manage/s/article/K000133566
Sounds like you may want to enable debug mode (https://my.f5.com/manage/s/article/K19451442) and observe what is happening. Your question is theoretical and may not be easily answered here, nor if you submit an F5 support case (this is like an engineering or product manager type of question). However, these resources shared here should give you enough tools to test and figure out the answer on your own accord.
