Forum Discussion

The_Mentalist_1's avatar
The_Mentalist_1
Icon for Nimbostratus rankNimbostratus
Oct 19, 2011

JsessionID cookie flush persist table

Hi guys

 

 

 

i use JsessionID iRule ,

 

 

 

but when a pool member is down , F5 take a few second (8-9) to redirect to other server.

 

 

 

i have a monitor configure in 2-3 second......

 

 

 

i think f5 take this time because in the persist table still appears the record of the pool member down.

 

 

 

so

 

 

 

How i can delete this record whit the iRule.?

 

 

 

thanks!

 

 

 

application delivery
dev
devops
iRules
  • Michael_Yates's avatar
    Michael_Yates
    Icon for Nimbostratus rankNimbostratus
    Oct 19, 2011
    The delay that you are seeing should just be the Timeout Value Time in the Health Monitor.

     

     

    Interval: 5 seconds

     

    Timeout: 16 seconds

     

     

    The Health monitor has to timeout before it will mark the node down.

     

     

    If you are wanting to rush things you could add another event in your iRule LB_FAILED Event and then use LB::down, persist none, and LB::reslect

     

     

    http://devcentral.f5.com/wiki/iRules.LB__down.ashx

     

    http://devcentral.f5.com/wiki/iRules.persist.ashx

     

    http://devcentral.f5.com/wiki/iRules.LB__reselect.ashx

     

     

    Hope this helps.
  • nitass's avatar
    nitass
    Icon for Employee rankEmployee
    Oct 21, 2011
    is irule logging helpful to see what is going on?

    e.g. 

    [root@iris:Active] config  b virtual bar list
virtual bar {
   snat automap
   pool foo
   destination 172.28.17.33:http
   ip protocol tcp
   rules myrule
   profiles {
      http {}
      tcp {}
   }
}
[root@iris:Active] config  b pool foo list
pool foo {
   members {
      10.10.70.110:http {}
      74.125.235.50:http {}
   }
}
[root@iris:Active] config  b rule myrule list
rule myrule {
   when CLIENT_ACCEPTED {
        log local0. "[IP::client_addr]:[TCP::client_port]"
}

when HTTP_REQUEST {
        log local0. "[IP::client_addr]:[TCP::client_port]|[HTTP::uri]"
}

when LB_SELECTED {
        log local0. "[IP::client_addr]:[TCP::client_port]|[HTTP::uri]|[LB::server addr]"
}

when LB_FAILED {
        log local0. "[IP::client_addr]:[TCP::client_port]|[HTTP::uri]|[LB::server addr]"

        LB::down
        LB::reselect
}

when HTTP_RESPONSE {
        log local0. "[IP::client_addr]:[TCP::client_port]|[IP::remote_addr]:[TCP::remote_port]"
}
}

Oct 21 18:25:45 local/tmm info tmm[17220]: Rule myrule : 172.28.17.30:56297
Oct 21 18:25:45 local/tmm info tmm[17220]: Rule myrule : 172.28.17.30:56297|/
Oct 21 18:25:45 local/tmm info tmm[17220]: Rule myrule : 172.28.17.30:56297|/|10.10.70.110
Oct 21 18:25:45 local/tmm info tmm[17220]: Rule myrule : 172.28.17.30:56297|/|10.10.70.110
Oct 21 18:25:45 local/tmm info tmm[17220]: Rule myrule : 172.28.17.30:56297|/|74.125.235.50
Oct 21 18:25:45 local/iris err mcpd[15514]: 01070809:3: iRule has marked pool member 10.10.70.110:80 down, no monitor to mark up.
Oct 21 18:25:45 local/iris notice mcpd[15514]: 01070638:5: Pool member 10.10.70.110:80 monitor status iRule down.
Oct 21 18:25:45 local/tmm info tmm[17220]: Rule myrule : 172.28.17.30:56297|74.125.235.50:80