Just noting this here for the record. If it happens to anyone else or
again to me I will try to get enough details for a good bug report. The
effect has since disappeared and the relevant Squid started working
properly.
** The apparent action to fix was setting debug_options ALL,9 (!!).
On upgrading a Squid 3.HEAD built 2013-12-16 to one built 2014-01-06
this strange HIT behaviour began to happen:
This cachemgr traffic used to be a MISS ...
GET /squid-internal-mgr/info HTTP/1.1
Host: treenet.co.nz
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:26.0)
Gecko/20100101 Firefox/26.0
Accept: */*
Accept-Language: en-gb,en;q=0.5
Accept-Encoding: gzip, deflate
X-Requested-With: XMLHttpRequest
Authorization: Basic *****
Connection: keep-alive
If-Modified-Since: Tue, 07 Jan 2014 01:03:44 GMT
HTTP/1.1 200 OK
Server: squid/3.5.0.0-0~fischeri20140106-0
Mime-Version: 1.0
Date: Tue, 07 Jan 2014 01:03:44 GMT
Content-Type: text/yaml
Expires: Tue, 07 Jan 2014 01:03:44 GMT
Last-Modified: Tue, 07 Jan 2014 01:03:44 GMT
Age: 1389062245
Warning: 113 treenet.co.nz (squid/3.5.0.0-0~fischeri20140106-0)
This cache hit is still fresh and more than 1 day old
X-Cache: HIT from treenet.co.nz
X-Cache-Lookup: HIT from treenet.co.nz:8080
Transfer-Encoding: chunked
Via: 1.1 treenet.co.nz (squid/3.5.0.0-0~fischeri20140106-0)
Connection: keep-alive
Problem #1:
It should be a MISS since the response is generated by the Squid
serving it up new on every request.
Problem #2:
see Warning header message statement vs Expires header.
Problem #3:
Age header is ~44 years
Amos
Received on Tue Jan 07 2014 - 04:58:13 MST
This archive was generated by hypermail 2.2.0 : Tue Jan 07 2014 - 12:00:10 MST