Connect_retry in 1.1.21

Hello All,

I have a Linux box running squid on accelerator mode with ipfwadm
redirection to take care of transparent proxying. Overall everything seems
to be OK so far, i.e. it re-directs all www requests etc, etc.

But I am regaulary getting the following types of errors in
/var/adm/messages for the named daemon.

May 21 03:59:57 plato named[94]: Lame server on
'www.{some-name}.com.global-bd.net' (in 'global-bd.net'?):
[202.76.120.3].53 'TXGROUP.NET'

Global-bd.net is our domain as well as txgroup.net. Of course this type of
lookup returns a server not found message to the client.
How does a site URL get appended with our domain name in a dns lookup by
Squid and why?
I thought it might be a BIND problem so I upgraded my BIND server to 8.1.2,
but still getting the same results.
I'm still not sure if there is something amiss in the BIND configuration.

But strangely, the same site that gets treated this way is handled normally
at most other times. Thats the strange part, this error appears
inconsistent. Would appreciate any pointers on what to do.

Just a related question, is the connect-retry fix implemented in Squid
1.1.21? If not, would this fix the problem mentioned above? How would I
patch it into 1.1.21?

Thanks in advance.
SFQ
GOS
Received on Wed May 20 1998 - 21:49:53 MDT