I noticed that squids use of filedescriptors is ever increasing.
Today, after a restart, the numbers would drop again - to normal levels.
I found that the decrease in filedescriptors (cacheCurrentFileDescrCnt)
correlated in a similar decrease in the number of clients
(cacheClients).
Coulod it be
Hi,
is it possible to re direct the requests coming at gateway from internet
in following format
rtmp://site1.mydomain.com
rtmp://site2.mydomain.com
rtmp://site3.mydomain.com
rtmp://site4.mydomain.com
by squid to internal servers which are actually hosting site1.mydomain.com
File Descriptor count and Cache Client count are not the same thing.
File Descriptors are needed whenever Squid has to write / read to a
file or the network (server or client side). So, even if only one
user is using, Squid might have needed 50 FDs.
If the cache.log is giving warnings about
* Nyamul Hassan mnhas...@usa.net:
File Descriptor count and Cache Client count are not the same thing.
File Descriptors are needed whenever Squid has to write / read to a
file or the network (server or client side). So, even if only one
user is using, Squid might have needed 50 FDs.
Yes.
I'm currently running 2.7.STABLE9, and don't see any FD leaking issue.
It might be in your version, but not common.
Please check the CacheMgr Info page, and the FD page for details on
how those FDs are being used. Perhaps it will make things more clear.
Our Squid has been running with the
* Nyamul Hassan mnhas...@usa.net:
I'm currently running 2.7.STABLE9, and don't see any FD leaking issue.
Yes, in 2.7 I also didn't have those problems :)
It might be in your version, but not common.
Please check the CacheMgr Info page, and the FD page for details on
how those FDs are
Hi, Henrik
On NTLM authentication environment, we could not do POST request
which size is over SQUID_TCP_SO_RCVBUF.
Should work. If not then file a bug report.
Have the problem already been fixed ? If still not, I'd like to do a bug report.
Sincerely,
--
Mikio Kishi
2010/6/7 Henrik
Hi, all
If the icap server(reqmod) responses the following,
HTTP/1.0 407 Proxy Authentication Required
Proxy-Authenticate: Basic realm=X
squid try to delete the Proxy-Authenticate header
client_side_reply.cc
1218 // if there is not configured a peer proxy with login=PASS
sön 2010-08-15 klockan 00:59 +0900 skrev Mikio Kishi:
Hi, Henrik
On NTLM authentication environment, we could not do POST request
which size is over SQUID_TCP_SO_RCVBUF.
Should work. If not then file a bug report.
Have the problem already been fixed ? If still not, I'd like to do a
Hello squid-users,
I'm using Squid in a reverse proxy setup. Recently I was surprised to
see some errors being served on popular pages since I have max_stale 1
day and stale-while-refresh=1200 set. Upon inspection I found Squid to
be replacing stale content (which had a return status of 200) with
Hi,
all on my brige+ squid + tproxy i'm running a web server.
and i cannot access to web page hosted on the box.
in my access.log i have a TCP_MISS/503 3157 GET
my config
+
iptables -t mangle -A PREROUTING -p tcp --dport 80 -j TPROXY
--tproxy-mark 0x1/0x1 --on-port
Mikio Kishi wrote:
Hi, all
If the icap server(reqmod) responses the following,
HTTP/1.0 407 Proxy Authentication Required
Proxy-Authenticate: Basic realm=X
squid try to delete the Proxy-Authenticate header
client_side_reply.cc
1218 // if there is not configured a peer proxy
Mamadou Touré wrote:
Hi,
all on my brige+ squid + tproxy i'm running a web server.
and i cannot access to web page hosted on the box.
in my access.log i have a TCP_MISS/503 3157 GET
my config
+
iptables -t mangle -A PREROUTING -p tcp --dport 80 -j TPROXY
13 matches
Mail list logo