Forum Discussion

mr_shaggy_17493's avatar
mr_shaggy_17493
Icon for Nimbostratus rankNimbostratus
Feb 26, 2018

Local Traffic Adaptive Monitoring

Hi All,

 

Anyone ever try to used adaptive monitoring on F5? As mention on LTM monitor reference document:

 

"A monitor with adaptive response time enabled marks a service as up or down based on the deviation of latency of the monitor probe from the mean latency of a monitor probe for that service. In typical cases, if the monitor detects three consecutive probes that miss the latency value you set, the system marks the pool member or node as down."

 

In my case, I'm going to use this adaptive monitor for monitoring ISP link on F5 Link Controller, run using BIGIP OS 13.1.0.2. I make a monitor object like this:

 

  • Type: gateway_icmp
  • Interval time: 5 sec
  • Timeout time: 16 sec
  • Transparent: Yes
  • Alias address: 8.8.8.8
  • Alias port: all ports
  • Adaptive: Enabled
  • Allowed Divergence: Absolute, 5 msec
  • Adaptive Limit: 5 msec
  • Sampling timespan: 60 sec

If I do ping test from CLI to 8.8.8.8, then I get the average latency time is 13 msec:

 

  • [root@CDT:Active:Standalone] config ping 8.8.8.8
  • PING 8.8.8.8 (8.8.8.8) 56(84) bytes of data.
  • 64 bytes from 8.8.8.8: icmp_seq=1 ttl=61 time=13.2 ms
  • 64 bytes from 8.8.8.8: icmp_seq=2 ttl=61 time=13.0 ms
  • 64 bytes from 8.8.8.8: icmp_seq=3 ttl=61 time=13.4 ms

I was intentionally set the allowed divergence and time limit to a low number, which is 5 msec, to see how this monitor will marked down the pool member, which in this case is an ISP link. But I never get this pool member to be marked down by this monitor object.

 

Anything wrong on my configuration? Do I need additional configuration to make it work?

 

Thanks anyway..

 

2 Replies

  • I'm having the same problem with a similar monitor; when I set the allowed divergence to 1ms, I still cannot get pool member to fail.

    Did you ever come to a resolution? I've tried this on 12.1.3.7 and 13.1.1.5 with the same results.

  • I'm having the same problem with a DNS/gateway ICMP monitor; 

    The latency show in tmsh show net monitor .... is 4ms. However, the pool is still up and not increase for the probe failure counter.

    Appreciate if any have solution.