EEM randomly stop the execution of differents actions no matter what level?

[josedunet]

Hello All,

I have configured ipsla + EEM to run an actions list when a node is down or UP.Below the configuration.

-------------------------------------------------------------------------------------------

ip sla monitor 10

type echo protocol ipIcmpEcho 10.18.48.1

timeout 500

frequency 15

ip sla monitor schedule 10 life forever start-time now

event manager applet SLA-DOWN

event snmp oid 1.3.6.1.4.1.9.9.42.1.2.9.1.6.10 get-type exact entry-op eq entry-val 1 exit-op eq exit-val 2 poll-interval 10

action 1.1 cli command "enable"

action 1.2 cli command "conf t"

action 1.3 cli command "interface fa0/0"

action 1.4 cli command "ip access-group MYTRAFIC out"

action 1.5 cli command "end"

action 1.6 cli command "exi"

event manager applet SLA-UP

event snmp oid 1.3.6.1.4.1.9.9.42.1.2.9.1.6.10 get-type exact entry-op eq entry-val 2 exit-op eq exit-val 1 poll-interval 10

action 1.1 cli command "enable"

action 1.2 cli command "configure terminal"

action 1.3 cli command "interface fa0/0"

action 1.4 cli command "no ip access-group MYTRAFIC out"

action 1.5 cli command "end"

action 1.6 cli command "exit"

---------------------------------------------------------------------------------------

What I found is that the EEM actions is randomly execute. A few times, all actions will be execute well and everything works.  Another times, just a few actions (1,2,3,4, or 1,2,3,4,5 or 1,2,3, ...)   will run and cause the task does not work. The execution of different actions stop at any level

Please which can cause this behavior and how can I solve it? .

Thanks in advance.

Best Regards,

[Joe Clarke]

First, you do not need action 1.6.  You do not need to "exit" the CLI session.  Both policies can end with the action 1.5.  Second, it could be that the policy is hitting the 20 second maximum runtime and terminating without completing all of the actions.  This can happen if the device is using AAA and/or heavily loaded.  What you can do to confirm this is to configure "debug event manager action cli" and see why the policies are not executing to completion.  If the problem is the maximum runtime, you can extend this by adding "maxrun X" to the end of your event lines (where X is a number greater than 20).