Date: Wed, 27 Nov 1996 19:46:11 -0800 (PST) From: Jaye Mathisen <mrcpu@cdsnet.net> To: hackers@freebsd.org, current@freebsd.org Subject: Constant "Connection reset by peer". Message-ID: <Pine.NEB.3.95.961127193004.14820z-100000@mail.cdsnet.net>
next in thread | raw e-mail | index | archive | help
It will be difficult to document all the details, but I have a problem that's been plaguing my systems for a while, but until recently, didn't affect me personally. Now it does, and I'm on the hunt. :) X1 = BSD/OS 2.1 newsserver X2 = FreeBSD 3.0-current, supped as of 11/23. X3 = FreeBSD 2.1.5 dated sometime in late October. from -stable. X2 is a testbed newsserver talking to X1. X2's console from innd is full of "Connection reset by peer's" from innfeed's being sent to it. In the opposite direction, there's no apparent problems. X1 serves as my outside newsserver as well, but I don't see any of these errors in innd on it. X3 is a machine that servers a significant number of pop/email clients, and there is a steady stream of these "connection resets" as well. A client will click his "checkmail", and get connection timed out instantly, and a "reset by peer" message pops up on the console. He clicks again, and it goes right through and works properly. The qualcomm popd logs the error as well. I also get it occasionally from sendmail I have popped a Fluke LANMETER 682 on my lan, and let it run for hours, and it doesn't find anything obviously wrong. I am at a bit of a loss as to where to start. X1, X2, and X3 are all on the same segment of the LAN, consecutive IP's. I get no other errors on the console, all the cards seem to work just fine. I'm using sysconfig tcp defaults for those tcp sysctl variables, (Can't remember them off the top). I be stumpified. If anybody has any suggestions, I'm more than happy to dig around. Heck, I'll even pay for some help.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.NEB.3.95.961127193004.14820z-100000>