Services down

[jim.brown]

I have a CSS11150 that handles roughly 1000 services. The last few times I have configured new services for sites added to the switch, the service reports as "down" with an icmp keepalive, even though I can ping the IP address on the server. My config gile is about 100k, but when I check the disk I see only about 13mgs free on a 4 gig hard drive. Anyone got any idea what's up with this?

[pknoops]

Jim,

1000 services on a CSS 11150 is hitting the limits as far as services/content rules go. You may want to consider using a "keepalive script". This allows for more services this reducing the amount of keepalives going in/out of the CSS. To test the service you can go into debug mode (llama) and do an "icp probe" to the service itself.

[jim.brown]

Thank you very much for the reply. Looking it over, I appear to have approx. 750 services running to support approx. 325 content rules. Do you have any idea what the limit per switch might be? Also, any suggestions or links to learn about configuring a keepalive script would be appreciated very much. Lastly, why would a 100k config file use up 4 gigs of hardrive space on the CSS?

[pknoops]

The limit is not really a hard fast limit, but in general we have always said 500 combined services and content rules. The catch with this is the types of keepalives and the frequency by which the keepalives are sent..

Here is a link for global keepalive scripts on out cco site:

http://www.cisco.com/univercd/cc/td/doc/product/webscale/css/css_610/bsccfggd/services.htm

As far as the hard drive space goes, I would check to see if you have any core dump files which can be as large as 256MB a piece. You can do a "show core" on the CSS to see if there are any. The only other thing on the CSS that can use a lot of space is the sys.log file.

Regards

Pete..

[jim.brown]

Thanks, Pete. I appreciate your time and help.

Jim