By the help of God. Hi, This is basically the network topology that I'm using: adsl <--> vrf <--> <eth1> [squid/icap machine] <eth0> <--> vrf <--> <internet>
When traffic goes via squid I see that eth1 (The one closes to adsl users) is very high this is from sar output: Average: IFACE rxpck/s txpck/s rxkB/s txkB/s rxcmp/s txcmp/s rxmcst/s %ifutil Average: lo 4191.83 4191.83 16976.00 16976.00 0.00 0.00 0.00 0.00 Average: eth0 2921.48 1224.57 3432.46 485.77 0.00 0.00 0.00 0.28 Average: eth1 7558.70 11544.74 920.91 *14447.44* 0.00 0.00 0.00 1.18 When traffic doesn't go via squid this is the output: Average: IFACE rxpck/s txpck/s rxkB/s txkB/s rxcmp/s txcmp/s rxmcst/s %ifutil Average: lo 19.10 19.10 2.25 2.25 0.00 0.00 0.00 0.00 Average: eth0 3666.40 2133.70 4608.70 409.59 0.00 0.00 0.00 0.38 Average: eth1 2213.40 3741.10 424.38 *4613.08* 0.00 0.00 0.00 0.38 I'm can't tell for sure that this is related but I saw several times the kernel prints: TCP: out of memory -- consider tuning tcp_mem The squid version I'm using is: /usr/local/squid/sbin/squid -v Squid Cache: Version 6.5-VCS Service Name: squid This binary uses OpenSSL 3.0.2 15 Mar 2022. configure options: '--with-large-files' '--with-openssl' '--enable-ssl' '--enable-ssl-crtd' '--enable-icap-client' '--enable-linux-netfilter' '--disable-ident-lookups' And I turned off persistence from client, icap and server sessions. What could be the problem? Thanks, Ben
_______________________________________________ squid-users mailing list squid-users@lists.squid-cache.org https://lists.squid-cache.org/listinfo/squid-users
