Critical Services on CSS - All or Any?

[glncalence]

I've configured vip redundancy on a pair of 11506. I would like critical services defined to failover when ALL the services are down or the upstream gateway is down. The documentation isn't very clear on this. There are simple examples and only the one line about an ALL SERVICES option. (see below)

"Local critical services for any service other than scripted or redundancy

uplink, such as a Web service. The VR goes down when all associated local

critical services go down."

I assume you do the ALL services option by creating multipe critical services under the IP Interface. But the documentation doesn't say. Am I right?

[Gilles Dufour]

this is explained at http://www.cisco.com/en/US/products/hw/contnetw/ps792/products_configuration_guide_chapter09186a008040b568.html#wp1091160

The "local critical service" is any service not using script keepalive or not configured as "type redundancy-up".

You are correct about your assumption.

You add all the critical services under the circuit.

Gilles.

[glncalence]

Hi Gilles,

Thanks for the weekend reply! I'd read that section on critical services. It lists three options. The first two will fail if ANY of the configured keepalives fail while the third we clarified how to do an ALL option.

What I really want to do is fail over when the upstream gateway is unreachable OR all services are down (basically a hybrid ANY/ALL option).

Rather than build a separate service should I use the existing ones? I'm already using port specific keepalives under my services (port 80 and a custom application TCP port).

Can I add a second keepalive type to each service with the ping script listing the server IP and upstream router IP? That way if the router fails the service is down.

I'm guessing if it isn't possible to add a second keepalive to a service I will need to write a custom script to accomplish what I'm trying to do. No biggie I'm just looking for the quickest way to do this.

Thanks in advance.