HI Guys,

I am seeing these lines in my cache.log:

2004/05/21 20:14:13| Config 'request_header_max_size'= 10240 bytes.
2004/05/21 20:15:26| Request header is too large (12287 bytes)
2004/05/21 20:15:26| Config 'request_header_max_size'= 10240 bytes.
2004/05/21 20:16:56| Request header is too large (12287 bytes)
2004/05/21 20:16:56| Config 'request_header_max_size'= 10240 bytes.
2004/05/21 20:16:58| Request header is too large (12287 bytes)

I checked my squid.conf, i set request_header_max_size = 10240 bytes still
i see this error in my cache.log. i searched the google and find the same
problem on the list
(http://www.mail-archive.com/[EMAIL PROTECTED]/msg16326.html)
but there is no clear solution for it.
I saw my access.log and find these line on it:

217.68.25.212 - - [21/May/2004:20:21:16 +0430] "NONE
error:request-too-large HTTP/0.0" 413 1592 NONE:NONE
217.68.25.212 - - [21/May/2004:20:21:16 +0430] "NONE
error:request-too-large HTTP/0.0" 413 1592 NONE:NONE

It likes a DOS attack or a virus, and it use many file descriptor so
sometimes these lines apear in cache.log

2004/05/21 20:23:00| WARNING! Your cache is running out of filedescriptors
2004/05/21 20:23:24| WARNING! Your cache is running out of filedescriptors
2004/05/21 20:23:49| WARNING! Your cache is running out of filedescriptors

I recompiled my squid with 32768 filedescriptors but the problem still
exist and it affects the performance very much.

Any help or suggestion highly appreciated.
Alireza

Reply via email to