Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!zaphod.mps.ohio-state.edu!samsung!uunet!mcsun!ukc!axion!delluk!tim From: tim@delluk.uucp (Tim Wright) Newsgroups: comp.unix.sysv386 Subject: Re: CLOSED TCP connection in Interactive 2.2 Message-ID: Date: 13 Dec 90 12:22:10 GMT References: <1990Dec11.203017.7862@mks.com> Sender: usenet@delluk.uucp (Usenet posting login) Organization: Dell Computer Corp., Bracknell, UK Lines: 29 In <1990Dec11.203017.7862@mks.com> eric@mks.com (Eric Gisin) writes: >We frequently get CLOSED TCP connections in Interactive UNIX 2.2. >For example, >tcp 0 0 femto.login mks.1021 CLOSED >is the result of a dead rlogin from mks to femto. >Once such a connection exists, it does not go away. >And if 1021 is the next available reserved port on mks, >"rlogin femto" will fail with "Connection timed out". >You can work around this by running two rlogins, the second will work. >It also happens with rsh, which is more serious for us because we >use rsh to do backups over the net. >Do others encounter this bug? I think 2.0 had this bug, and the 2.2.1 >summary doesn't mention this bug being fixed. Is there any way to kill this >CLOSED connection without rebooting? Yes I've seen this. If a PC is telnetted in to the system and is switched off, the telnetd on the host side does not die. The annoying part is that this *did* work with 1.1.2. I don't know what's broken, KEEPALIVE code ? TCP timers ?. Somebody running SCO UNIX told me that they have the same problem with TCP/IP there. Can anybody (Lachman ?) comment ? Tim -- Tim Wright, Dell Computer Corp. (UK) | Email address Bracknell, Berkshire, RG12 1RW | Domain: tim@dell.co.uk Tel: +44-344-860456 | Uucp: ...!ukc!delluk!tim "What's the problem? You've got an IQ of six thousand, haven't you?"