Re: [squid-users] comm_open: socket failure: (55) No buffer space available & Limiting closed port RST response from 201 to 200 packets/sec

From: Adrian Chadd <adrian@dont-contact.us>
Date: Fri, 10 Nov 2006 22:31:55 +0800

On Fri, Nov 10, 2006, Tek Bahadur Limbu wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
>
> Dear All,
>
> I am facing some problems on my FreeBSD 6.0 squid box.
>
> I am using Squid-2.6.Stable5 running on FreeBSD 6.0 i386 on a Dell
> SC420.
>
> I get the following errors on my cache.log.
>
>
> 2006/11/10 16:31:06| comm_open: socket failure: (55) No buffer space
> available
>
> 2006/11/10 16:31:06| comm_open: socket failure: (55) No buffer space
> available
> 2006/11/10 16:31:06| comm_open: socket failure: (55) No buffer space
> available2006/11/10 16:31:07| parseHttpRequest: Unsupported method 'aa
>
> 2006/11/10 16:31:07| clientReadRequest: FD 2695 (x.x.x.x:21026)
> Invalid Request
> 2006/11/10 16:31:08| parseHttpRequest: Unsupported method 'REGISTER'
>
>
> When I look at my /var/log/message, I get the following:
>
> Nov 10 16:31:21 proxy6 kernel: Limiting closed port RST response from
> 232 to 200 packets/sec
> Nov 10 16:35:34 proxy6 kernel: Limiting closed port RST response from
> 201 to 200 packets/sec
>
> Nov 10 16:36:17 proxy6 last message repeated 6 times
>
> I guess that that the first problem is related to networking related
> buffers. But what does the 2nd error mean?

Take a look at the output of netstat -mb:

nympho# netstat -mb
77/144/131072 mbufs in use (current/peak/max):
        65 mbufs allocated to data
        12 mbufs allocated to packet headers
64/80/32768 mbuf clusters in use (current/peak/max)
196 Kbytes allocated to network (0% of mb_map in use)
0 requests for memory denied
0 requests for memory delayed
0 calls to protocol drain routines

See if you've run out of membufs. FreeBSD versions prior to 6 used to log
something in dmesg when it was running out of nmbclusters; I wonder why
it stopped.

It lmight also be a full accept queue for incoming connections.

sysctl kern.ipc.somaxconn ; might need to up it a little.

Second is the TCP limiting stuff.

sysctl -w net.inet.icmp.icmplim=0

Here's /etc/sysctl.conf on my FreeBSD test server. I'd appreciate feedback
if I've boneheaded anything:

nympho# cat /etc/sysctl.conf
net.inet.tcp.sendspace=65536
net.inet.tcp.recvspace=65536
vm.defer_swapspace_pageouts=1
net.inet.tcp.msl=3000
kern.maxfiles=32768
kern.maxfilesperproc=16384
net.inet.icmp.icmplim=0
kern.ipc.shmseg=128
kern.ipc.shmall=16384
net.inet.tcp.delayed_ack=0

and /boot/loader.conf:

nympho# cat /boot/loader.conf
# -- sysinstall generated deltas -- #
userconfig_script_load="YES"
kern.ipc.maxsockets=32768
kern.ipc.nmbclusters=32768
kern.ipc.nmbufs=131072

Adrian
Received on Fri Nov 10 2006 - 07:27:20 MST

This archive was generated by hypermail pre-2.1.9 : Fri Dec 01 2006 - 12:00:03 MST