Educause Security Discussion mailing list archives

Re: Squid


From: Kevin Halgren <kevin.halgren () WASHBURN EDU>
Date: Mon, 27 Jun 2011 09:23:29 -0500

Using cachemgr.cgi (has to be moved into Apache cgi-bin directory, fyi), under "General Runtime Information" and looking at the Byte hit ratios. Since the primary purpose was to server as a proxy, monitoring the caching stats long-term hasn't been a priority. However, spot-checking under typical use we see around 10%

Kevin

On 6/24/2011 12:46 PM, Kevin Wilcox wrote:
On Fri, Jun 24, 2011 at 9:13 AM, Kevin Halgren
<kevin.halgren () washburn edu>  wrote:

We've retained the
caching function because it's helpful when downloading large updates to
refresh systems, nonetheless we've only seen about a 10% reduction in
traffic from utilizing the squid cache.  At that level, installing a squid
proxy for general campus use simply isn't worth the maintenance and
occasional trouble tickets that it would generate.
How are you measuring the saved bandwidth? SNMP against the squid servers?

kmw


--
Kevin Halgren
Assistant Director - Systems and Network Services
Washburn University
(785) 670-2341
kevin.halgren () washburn edu


Current thread: