ernie@nucsrl.UUCP (Ernest Woodward) (10/15/87)
I am once again requesting information concerning the operational use of two DEQNA's and Mt. Xinu. I have posted a related question in the past with less description of the problem and received some useful but incomplete replies. I will try to fully describe the situation. Following is the console display at boot-up (into single-user mode): Boot : ra(0,0)vmunix 260936+90516+65580 start 0x107c 4.3 BSD UNIX #0: Tue Oct 13 08:10:01 CDT 1987 ernie@accuvax.nwu.edu:/usr/sys/accuvax real mem = 9433088 avail mem = 8022016 using 281 buffers containing 576512 bytes of memory uba0 at tr0 uda0 at uba0 csr 172150 vec 774, ipl 15 ra0 at uda0 slave 0 tmscp0 at uba0 csr 174500 vec 770, ipl 17 tms0 at tmscp0 slave 0 zs0 at uba0 csr 172520 vec 224, ipl 15 ts0 at zs0 slave 0 qe0 at uba0 csr 174440 vec 764, ipl 17 qe1 at uba0 csr 174460 vec 760, ipl 17 dhu0 at uba0 csr 160440 vec 300, ipl 17 dda0 at uba0 csr 167000 vec 750, ipl 15 # For a quick description of the system, I have a MicroVAXII with 9M of memory, 2 DEQNA's, TS05 tape drive, TK50 tape drive, ACC's ACP5250, one DHV11, KDA50 disk controller with one RA81 drive, and all is enclosed in the large cabinet with the two BA23 backplanes. I had CE's run power tests on the two backplanes and they claim there is no problem with power supply. Once I have the shell in single-user mode, I manually enter the following commands: #/etc/mount /dev/ra0g /usr #/etc/mount /dev/ra0h /usr/users #ifconfig qe0 inet 192.26.85.1 broadcast 192.26.85.0 -trailers #ifconfig qe1 inet 128.174.5.131 broadcast 128.174.0.0 -trailers #ifconfig lo0 127.0.0.1 # I run netstat -in to check if the two DEQNA's are running and then I enter the CTRL-D to start up multi-user mode. I run with the very simple /etc/rc file for testing this system problem. Below is a copy of the file: hostname accuvax.nwu.edu hostid 'hostname' ps -ax >/dev/console date >/dev/console exit 0 Once I have entered multi-user mode, the system seems to get paralyzed. The login prompts are displayed and sometimes I can enter a login name and password and receive /etc/motd, but I never get the shell prompt. I will also make three quick points concerning the use of two DEQNA's and Mt Xinu that may be useful for understanding the issues. (1) Some people have recommended various combinations of ifconfig statements with the two DEQNA's and the local loopback. I have tried all combinations of starting the network via ifconfig and the multi-user mode environment will freeze at a rather unpredictable point in time. (2) We have been able to use the MicroVAXII and Mt. Xinu for several days while networking with one DEQNA and the ACP5250. We have also been able to run the machine as a gateway for two DEQNA's and the ACP5250 while operating in single-user mode. We have been able to freeze the shell for single-user mode when rebuiling a kernel and operating as a gateway. (3) We have a simple monitor program which displays the process table and run queue when activated at the console. It appears that processes are not in the run queue but are listed in the process table as running (during the paralyzed state described above). Do any wizards have a clue as to what the problem may be? Is it a bug in the kernel (handling process scheduling), the UDA driver, or in ifconfig? Does anyone have an idea of where I should start? Are there newer versions of the UDA or DEQNA driver for the Mt. Xinu system? I would greatly appreciate anyones help. I have been working on this problem for 6 weeks and I know relatively very little UNIX. Please post notes or send mail to: ernie@eecs.nwu.edu Thanks Sincerely, Ernie Woodward Academic Computing and Network Services Northwestern University