Path: csiph.com!v102.xanadu-bbs.net!xanadu-bbs.net!feeder.erje.net!us.feeder.erje.net!newsfeed.fsmpi.rwth-aachen.de!news-1.dfn.de!news.dfn.de!fu-berlin.de!uni-berlin.de!individual.net!not-for-mail From: Tim Watts Newsgroups: comp.os.linux.misc Subject: Re: weirdness with TCP timeouts? Date: Mon, 08 Sep 2014 17:35:01 +0100 Lines: 23 Message-ID: References: Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit X-Trace: individual.net RkMfpSCqh/aXZM9T5Rlu4gllR1lkC/PYdM6Zoy3V3L3NCp+y68 X-Orig-Path: squidward.local.dionic.net!not-for-mail Cancel-Lock: sha1:tfyJaMOCzQdU1jfLsnnY4U9BEtU= User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:24.0) Gecko/20100101 Thunderbird/24.2.0 In-Reply-To: Xref: csiph.com comp.os.linux.misc:12026 On 08/09/14 11:25, The Natural Philosopher wrote: > The second issue was that ssh sessions to my remote server were timing > out. They never used to, but I pulled the ssh keepalive down to a couple > of minutes and now the session stays up. As that is possibly a reproducible and simple problem, could you put the timeout up and run a tcpdump against it until it dies and see if anything happens at the end? > That is in some way sockets are being closed inside the kernel in ways > they shouldn't be. And applications are not being notified that they > have closed. > > IN short my fevered imagination is considering the possibility that TCP > timeouts in the kernel have changed in unexpected ways, so that > connections that used to exist, have been silently dropped. Not aware of anyhing which means bugger all here - Just one observation - have you changed the device that provides your NAT (if any) to the internet recently? NAT clearing sessions could cause these symptoms