Cisco Support Community
cancel
Showing results for 
Search instead for 
Did you mean: 
New Member

Request clarification on serverfarm/rserver/probe behaviour

Hi,

  I'm working with my networking team trying to get our loadbalancer/failover configured.  I've spent quite a bit of time pouring over the ACE documentation, and I'm still unclear on the effect of a failed probe when the probe is applied to an rserver in a serverfarm.  In the (simplified) example below, each rserver has an http server that sits in front of two servlet containers providing the content at urls /app1 and /app2.

The question: If (only) PROBE-APP-1 failed against HTTP-1 in SF-APP-1, would HTTP-1 be considered "down" within SF-APP-2?

Thanks in advance!

rserver host HTTP-1
  ip address 10.0.1.100

rserver host HTTP-2
  ip address 10.0.1.101

probe http PROBE-APP-1
  request method head url /app1/probe
  expect status 200 200

probe http PROBE-APP-2
  request method head url /app2/probe
  expect status 200 200

serverfarm SF-APP-1
  rserver HTTP-1
    probe PROBE-APP-1
    inservice
  rserver HTTP-2
    probe PROBE-APP-1
    inservice

serverfarm SF-APP-2
  rserver HTTP-1
    probe PROBE-APP-2
    inservice
  rserver HTTP-2
    probe PROBE-APP-2
    inservice

edited for formatting...

1 ACCEPTED SOLUTION

Accepted Solutions
Cisco Employee

Request clarification on serverfarm/rserver/probe behaviour

Hi Brent,

If (only) PROBE-APP-1 failed against HTTP-1 in SF-APP-1, would HTTP-1 be considered "down" within SF-APP-2 ?

Answer: No.

HTTP-1 be considered "down" within SF-APP-2 only when probe PROBE-APP-2 fails.

regards,

Ajay kumar.

2 REPLIES
Cisco Employee

Request clarification on serverfarm/rserver/probe behaviour

Hi Brent,

If (only) PROBE-APP-1 failed against HTTP-1 in SF-APP-1, would HTTP-1 be considered "down" within SF-APP-2 ?

Answer: No.

HTTP-1 be considered "down" within SF-APP-2 only when probe PROBE-APP-2 fails.

regards,

Ajay kumar.

New Member

Request clarification on serverfarm/rserver/probe behaviour

Thanks much for the answer.  I assumed that might be the behaviour, as it seems sensible, but was unable to feel confident of that assumption after reading over the documentation many times.

Is my understanding correct that HTTP-1 could be considered "active" (OPERATIONAL state) in SF-APP-2 at the same time it is considered "down" (PROBE-FAILED state) in SF-APP-1?


Cheers!

  Brent

257
Views
0
Helpful
2
Replies
CreatePlease to create content