Forum Discussion

Amr_Ali's avatar
May 09, 2023

node health check

I need to know why after I attached to my node health monitor type HTTP the node went down, I am sure the HTTP service is up on this server ( node ) and I noted this after attaching this node with pool and configuring HTTP health monitor on this pool and the pool member up without any issue,

so how after I attached the HTTP health monitor on the node appeared down and after attaching the same health monitor to pool that I attached to it the same health monitor the pool member ( node IP + port ) it appear up

1 Reply

  • Amr_Ali Typically health monitors such as HTTP have an undefined port number which is inherited from the pool members port number when the health monitor is performed. Most likely what is happening is that because you have an undefined port on the health monitor and the node has no associated port, the F5 does not know where to send the health monitor to so it fails. Typically you can click on the node and it will tell you why the health monitor is failing. In addition to this you typically do not assign a health monitor at the node level and you only have one assigned at the pool member level because in almost all instances the pool member health monitor passing means the node is healthy. If you did require a health monitor for the node and the pool member they typically would not be the same health monitor because that is redundant and both would return the same response. You should assign a different health monitor at the node level such as ICMP and then HTTP at the pool member level. Realistically though, you keep all health monitors off of the node level and just assign 1 or more at the pool member level and require all health monitors pass checks to consider a pool member alive. The node should just be an on/off switch for everything associated to it which you would use the enable/disable as the on off at the node level.