Path: utzoo!utgpu!water!watmath!uunet!lll-winken!ubvax!ames!killer!loci From: loci@killer.UUCP (loci!clb) Newsgroups: unix-pc.general Subject: Re: More silly 3b1 quesitons. :-) Summary: PANICS are just that Message-ID: <3994@killer.UUCP> Date: 5 May 88 19:34:56 GMT References: <431@jonlab.UUCP> <28@booboo.UUCP> Organization: The Unix(R) Connection, Dallas, Texas Lines: 34 From: clb@loci.uucp (CLBrunow) In article <28@booboo.UUCP>, spear@booboo.UUCP (Steve Spearman) writes: > .... but some people experienced system > panics that were believed to be related. You can certainly try the > program, but if you see unexplained panics you didn't have before, > you might want to discontinue use. > I don't believe that PANICS are any more rational in the 7300 than in people, but rather a loss of good sense. Panic traps have an amusing variety of messages and once my machine did something really bizzare: suddenly the top line (with the date, etc.) went sliding to the bottom of the screen, being followed by what looked like bars (vertical) so that the effect was like a post office or bank teller window slamming shut. I practically fell out of my chair, but rebooting made everything alright again. In fact, rebooting has always brought me back. I have a ritual that I have to go through about twice a year when the panics start happening frequently: I reformat the hard disk and reload the system and that helps a lot. My explanation is that the file system gets fragmented and the disk starts making seek errors as it jumps all over the place to read files (this is only my theory). Some years ago, I worked at Mostek where we tried to explain soft errors in RAM chips. The explanation for many of them was hits by ionizing radiation (cosmic rays) which can't be predicted or shielded, and that led to much discussion about large memories arrays (mega-bytes) and how the probability would increased for errors to the point that on-board error correction would be neccessary. Bottom Line: DON'T PANIC (you've heard that before, haven't you)