ACE configuration check-up!

syjeon · ‎04-22-2010

Hi everyone.

I would like to check the following ace configuration,

My main question is that the ace config like below is same function?

case.1

serverfarm host NAME_of_SERVERFARM

rserver REAL01

probe TCP50000

inservice

rserver REAL02

probe TCP50000

inservice

case.2

serverfarm host NAME_of_SERVERFARM

probe TCP50000

rserver REAL01

inservice

rserver REAL02

inservice

we forcus on probe,

in case,1 the probe is assigned on each real server, but in case2, the probe is only server farm, not on each rserver.

in this case. case1, case2 is same function? or not. Is there something special in two case?

finally!

in case.3

serverfarm host NAME_of_SERVERFARM

probe icmp

rserver REAL01

probe TCP50000

inservice

rserver REAL02

probe TCP50000

inservice

this case is in probe, on each rserver and server farm, on server farm check on icmp and rserver check the service port.

in that case, if just one probe is dead, the server farm declare the rserver is dead ,regardless of tcp is up.

Thank you.

ciscocsoc · ‎04-22-2010

Hi,

Case 1 and Case 2 are equivalent. A probe applied to a serverfarm is applied to all rservers within it.

Logically

serverfarm host NAME_of_SERVERFARM

probe icmp

rserver REAL01

probe TCP50000

inservice

rserver REAL02

is equivalent to:

serverfarm host NAME_of_SERVERFARM

probe icmp

rserver REAL01

probe TCP50000

probe icmp

inservice

rserver REAL02

probe TCP50000

probe icmp

inservice

so if either probe on an rserver fails it will make that rserver out of rotation.

This could be written as

serverfarm host NAME_of_SERVERFARM

probe icmp

probe TCP50000

rserver REAL01

inservice

rserver REAL02

inservice

If you want to have an rserver taken out of rotation only if all the probes on it fail then you need to add the "fail-on-all" command.

HTH

Cathy