lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening PHC | |
Open Source and information security mailing list archives
| ||
|
Date: Mon, 30 Nov 2009 23:19:39 +0100 From: Frederic Leroy <fredo@...rox.org> To: "Ilpo Järvinen" <ilpo.jarvinen@...sinki.fi> Cc: Netdev <netdev@...r.kernel.org>, Asdo <asdo@...ftmail.org> Subject: Re: scp stalls mysteriously Le Mon, 30 Nov 2009 23:37:56 +0200 (EET), "Ilpo Järvinen" <ilpo.jarvinen@...sinki.fi> a écrit : > Ok. Just wanted to confirm since it can be that that not all packets > are in the tcpdump that TCP was sending. For what reason is unknown > at the moment. Any fancy firewall (iptables) rules on houba that > could do such a trick here? Nothing with netfilter or schedulers on houba, neither on sg ( no log because no iproute installed ...) http://www.starox.org/pub/scp_stall/houba.netfilter.log > Also, have you changed sysctl settings from the defaults? No but I could have change defaults in the linux config file. I saved the current content of /proc/sys/net and proc/net : http://www.starox.org/pub/scp_stall/sg-proc_net-proc_sys_net.tbz2 http://www.starox.org/pub/scp_stall/houba-proc_net-proc_sys_net.tbz2 I put also linux config files in the directory. > Could you try btw to disable frto (set tcp_frto sysctl to zero), in > theory it could cause zero cwnd if something else is wrong too (I > haven't yet read them all through so there might be other places > where a zero cwnd can be set). Also I'm interested what happens if > you disable timestamps, is the problem still reproducable. Ok, I'll try these tomorrow when it will fault again :) -- Frédéric Leroy -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@...r.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists