Path: utzoo!utgpu!water!watmath!uunet!lll-winken!killer!osu-cis!att!westmark!argon!ebh From: ebh@argon.UUCP (Ed Horch) Newsgroups: unix-pc.general Subject: Re: 3.51a keeps dying; even HDB is a mess; any ideas out there? Message-ID: <289@argon.UUCP> Date: 9 Jul 88 03:29:17 GMT References: <422@kosman.UUCP> <1041@flatline.UUCP> <415@icus.UUCP> Reply-To: ebh@argon.UUCP (Ed Horch) Distribution: unix-pc Organization: Broke Broken Brokerage Lines: 26 Well, I'm running 3.5 (not 3.51[a]), and I experience a similar problem with the system locking up for several seconds at a time during heavy UUCP transfers via the OBM. However, this seems to be harmless, if a little annoying. I run a full newsfeed here (yes, at 1200 baud; I'm saving my pennies for a Trailblazer), and have never seen any harm done by this lockup problem. It also doesn't generate any messages in the unix.log file. It does seem to grab the *entire* system, though, since the curser even stops flashing. I can't believe that something that can grab the CPU for five to ten seconds at a time isn't a bug. Speaking of bugs.... Can somebody send me the fix to rn to get around the "Can't open /dev/tty" message when dialing in from the OBM? Also, is there any fix or workaround to the famous lost-inode problem that occurs most often during news unbatching? It's a real pain, more because of its seemingly complete randomness than anything else. Also, a reliability update: I have been running continuously for over seven months with exactly ZERO unexplainable crashes. I have a 7300 (two fans) with a 1MB motherboard, and a Miniscribe 6085 disk that sits outside the machine. -Ed