romwa@gpu.utcs.toronto.edu (Mark Dornfeld) (06/12/88)
System: SCO Xenix V 386 ESDI 2.2.2, Compaq 386 16Mhz, Omti 8620 ESDI controller, Compaq 130 Mb disk. A while back I posted a note about one of my systems which was reporting bad blocks on the hard disk but bad tracking found no flaws. I was able to trace the bad writes to a uucp process. All the bad writes occurred coincidentally with my hub uucp system calling this one over a 9600 bps direct line. This hinted that the disk wasn't bad, but rather something was out of sync with the uucp programs. A friend suggested I slow down the uucp and now I've gone a whole week without an error message. Seems to me like there is something out of whack in the timing of the uucp disk writes and the controller is sending an error message back to the kernel. I also did some rather confusing calculations to find out which cylinders/tracks the had the errors and the pattern I found was so different from any list of bad tracks I've seen that this strengthened my suspicion that the problem was not the disk. Has anyone else noticed this problem? SCO, do you know anything about this? Mark T. Dornfeld Royal Ontario Museum 100 Queens Park Toronto, Ontario, CANADA M5S 2C6 mark@utgpu!rom - or - romwa@utgpu