squid-2.3.STABLE1 dumps core with "http_port <ip_addr>:<port>"

From: Eelco M Glasl <emglasl@dont-contact.us>
Date: Thu, 17 Feb 2000 12:24:39 +0100 (MET)

Hi all,

my network configuration requires Squid in httpd_accel mode to be bound to
_one_ specific ip-address (out of more) per host, running Solaris 2.7.
So I used to configure squid-2.2.STABLE4 with:

http_port 80
tcp_incoming_address 192.168.1.1
tcp_outgoing_address 192.168.1.1
http_accel_host virtual
http_accel_port 80

This was working perfectly well. Until I tried to migrate to
squid-2.3.STABLE1 with the following configuration:

http_port 192.168.1.1:80
tcp_outgoing_address 192.168.1.1
http_accel_host virtual
http_accel_port 80

All that Squid wanted to do was dumping core with a segmentation fault.

Later I found out that Squid insisted on resolving the address 192.168.1.1
so I added a hostname for 192.168.1.1 in /etc/hosts, even if I didn't want
to use that name in "http_port" of squid.conf. I still use

http_port 192.168.1.1:80

... but now it works. Would you consider this behavior a bug?

Thanks for you input,
Eelco

-- 
/* Eelco M. Glasl <emglasl@gmx.net> */
Received on Thu Feb 17 2000 - 04:36:43 MST

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:51:19 MST