john@chinet.chi.il.us (John Mundt) (02/08/90)
I have a 3b1 which suddenly gets tied up after I call *into* it from another machine. When I log out with a ^D, the 3b1 does not disconnect but rather gives a "login" message. However, it never accepts a login, always saying "login incorrect." When I force a disconnect from outside, the 3b1 will no longer answer the line. The machine has been fine for two years before this, and even now, uucico can get in and out from the outside without tying up the line. Going to the machine shows no getty or login running on the phone port, even though /etc/inittab should respawn a getty. Doing a "init q" will respawn the getty, but the modem still does not answer the phone. The process table doesn't show any process, zombie or otherwise, that could be connected to the phone port. phtoggle does not clear the problem. Rebooting the machine is the only way to clear the problem. The hotline's suggestion is to reload UNIX. Has anyone else had this problem and a solution for it? -- --------------------- John Mundt Teachers' Aide, Inc. P.O. Box 1666 Highland Park, IL john@admctr.chi.il.us *OR* fred@teacha.chi.il.us (312) 998-5007 (Day voice) || -432-8860 (Answer Mach) && -432-5386 Modem
kls@ditka.UUCP (Karl Swartz) (02/09/90)
In article <1990Feb8.155753.3090@chinet.chi.il.us> john@chinet.chi.il.us (John Mundt) writes: >I have a 3b1 which suddenly gets tied up after I call *into* it from >another machine. When I log out with a ^D, the 3b1 does not disconnect >but rather gives a "login" message. What have you changed recently? Any chance you've changed stty settings in your .profile somewhere along the way, or something you've recently installed is doing this? Try 'stty -a'. If it says '-hupcl' instead of 'hupcl' that's probably why ^D doesn't disconnect you. >However, it never accepts a login, always saying "login incorrect." That's rather strange ... >Going to the machine shows no getty or login running on the phone >port, even though /etc/inittab should respawn a getty. A getty or login process on ph[01] shows up on terminal '?' for some reason, though on one of the serial ports they show up on 'tty00[0-6]' just fine. >Doing a "init q" will respawn the getty, but the modem still >does not answer the phone ... phtoggle does not clear the problem Try setgetty ph1 0 wait a few seconds, then setgetty ph1 1 See if that has any more success than phtoggle. Sometimes it will get something unstuck. Are you running the OBM on ph0? The software doesn't always seem to handle this correctly -- you're better off running it on ph1 even if you leave ph0 unused. (If you use UA you'll have to tell it there's voice line on ph0 aka LINE 1 to keep it happy.) What version of the software? Since you said getty and not uugetty I assume you aren't running HDB. >The hotline's suggestion is to reload UNIX. Has anyone else had >this problem and a solution for it? Figures. I've never gotten anything even resembling a helpful answer from them. "Six foot flames are shooting out the vents on my monitor!" "Have you tried reloading UNIX?" -- Karl Swartz |UUCP uunet!apple!zygot!ditka!kls 1-408/223-1308 |INet zygot!ditka!kls@apple.com "I never let my schooling get in |BIX kswartz the way of my education."(Twain) |Snail 1738 Deer Creek Ct., San Jose CA 95148
donlash@uncle.UUCP (Donald Lashomb) (02/13/90)
In article <1990Feb8.155753.3090@chinet.chi.il.us> john@chinet.chi.il.us (John Mundt) writes: >I have a 3b1 which suddenly gets tied up after I call *into* it from >another machine. When I log out with a ^D, the 3b1 does not disconnect >but rather gives a "login" message. However, it never accepts a >login, always saying "login incorrect." When I force a disconnect >from outside, the 3b1 will no longer answer the line. The machine has >been fine for two years before this, and even now, uucico can get in >and out from the outside without tying up the line. > >Going to the machine shows no getty or login running on the phone >port, even though /etc/inittab should respawn a getty. Doing a >"init q" will respawn the getty, but the modem still does not >answer the phone. > >The process table doesn't show any process, zombie or otherwise, >that could be connected to the phone port. phtoggle does not >clear the problem. Rebooting the machine is the only way to >clear the problem. > >The hotline's suggestion is to reload UNIX. Has anyone else had >this problem and a solution for it? >-- >--------------------- >John Mundt Teachers' Aide, Inc. P.O. Box 1666 Highland Park, IL >john@admctr.chi.il.us *OR* fred@teacha.chi.il.us >(312) 998-5007 (Day voice) || -432-8860 (Answer Mach) && -432-5386 Modem Yes, I am having a related problem with my 7300 since installing FixDisk2. Here is what I recently posted: ==================== Unix-pc group, +----------- has happenned 8 or 9 times now! v I have experienced 2 lockups since installing FixDisk2 of a type that has never happened to me before. It seems that init sometimes fails to respawn getty after I logout. I am pointing my finger at FixDisk2 because: 1. It never happened before to me. 2. FixDisk2 included new versions of init and getty. 3. There seems to be some mis-handling of wtmp/utmp, exactly the area of concern that FixDisk2 addressed. There is a remote chance that this could be a problem in wind.o, since of course, init is supposed to respawn getty /dev/window. ---------------------- Here are the symptoms: Welcome to the Unix PC Please login: install Password: $ windy -b ksh # I like having 2 shells [ do normal work everything seems ok ] $ <cntl-D> # exit second shell $ <cntl-D> # exit first (login) shell [ screen blanks - NO Welcome, NO Please login ] Point mouse at ph mngr window, click. Call screen of install appears - appearently ph thinks install is still logged in. This might be trouble with wtmp/utmp files. Get a shell so I can see what is going on :-> # ps -e [ no getty on w1 ] # init q [ no difference ] # killall # sync # reboot I don't think my playing around with having two shells going (windy -b ksh) has anything to do with the problem. I am careful to exit the second shell, thereby closing its window, before I exit the first shell (logout). I only include it here for completeness of the discussion. ---------------------- If anyone has had a similar experience, before or after installing FixDisk2, please post or email me what you know about it. Thanks in advance, -Don donlash@uncle.UUCP ==================== I still have not been able to find out what the problem is. I was able to confirm that getty is definately NOT being respawned by init. There seems to be some weird modification/access/stat-change times with utmp. I tried reloading the old version of init (keeping new versions of getty and login). This seemed to work for 4-5 days, then Wham, the machine locked up twice. The worse part of this problem is that it is so unpredicatable. I have yet to figure out just what circumstances it happens under. I don't think I'm doing anything other than what I always do and sometimes it happens, some- times it doesn't (race condition, missed signal, phase of the moon?). Currently, I'm running the old init, getty, and login, but the new 3.51m kernel. Haven't had any trouble so far, but then again it's only been half a day. I won't feel comfortable, until it's been at least a week with no trouble (is that long enough to be sure?). Oh yeah, I've double checked the sums of my fix disk with those posted on the net, there ok. I've even run memory and disk checks from the Diagnostic Disk - everything checks out ok. If anybody out there has any ideas as to what the problem is, *PLEASE* let me know. Thank you Lenny for your help so far. -Don donlash@uncle.UUCP