news@janhlbv.UUCP (news) (03/01/90)
We are having serious problems with 386/IX on a PS/2 model 80. We have 3 systems with the same configuration which are runing 24 hours a day. This systems are frequently crashing. We are getting the the following panic: - Panic: 0x0e This error is from a instabile TCP/IP and is acknowledge by interactive. 3 months ago and until no now still no bug fix! - Panic: Panic flags set on 17 (ESDI controler) track XXXX. - Some time the system just hangs. These systems are communication front ends for a mainframe and must by up 24 hours a day! Over the last 4 mounts we are having a average of 2 crashes a week. We have the following questions: - Is there anybody who is using PS/2 model 80 with UNIX? - Is there anybody using digiboard? - How stabile is TCP/IP (it seems very unstabile). The configuration is: - 386/IX version 2.0.2 - TCP/IP version 1.1.2 - PS/2 model 80-111 - 8 Mb of memory. - 2 Digiboard communication boards. - WD8003 Ethernet adapter.
haugen@bulus3.BMA.COM (John M. Haugen) (03/06/90)
In article <482@janhlbv.UUCP>, news@janhlbv.UUCP (news) writes: > We are having serious problems with 386/IX on a PS/2 model 80. > We have 3 systems with the same configuration which are runing 24 hours > a day. This systems are frequently crashing. We are getting the the following > panic: > - Panic: 0x0e > - Some time the system just hangs. Yes, I have seen a similar problem. > We have the following questions: > - Is there anybody who is using PS/2 model 80 with UNIX? > - Is there anybody using digiboard? > - How stabile is TCP/IP (it seems very unstabile). > > The configuration is: > - 386/IX version 2.0.2 > - TCP/IP version 1.1.2 > - PS/2 model 80-111 > - 8 Mb of memory. > - 2 Digiboard communication boards. > - WD8003 Ethernet adapter. I have a similar setup to you ( 386 based micro-channel machine with a WD8003 Ethernet adapter ) and had a similar problem. The problem turned out to be the Unix driver assumed the 8003 had 8K bytes of memory when it actually has 16K! You need modify the file in /etc/conf/pack.d/wd/space.c and change the defines for WDBOARDSIZE0 from 8192 to 16384. Then rebuild the kernal. Since making this fix, I have not had any more failures of that type. A better fix would be for the code to actually use the memory size specified in /etc/conf/sdevice.d/wd where the beginning and ending shared memory addresses are specified. Hope this fixes your problem. John M. Haugen Domain: haugen@BMA.COM Bull Micral of America UUCP: ...!uunet!bulus3!haugen 1970 Oakcrest Ave. #300 ATT: 612-633-5660 St. Paul, MN 55113-2624