lzm@mace.cc.purdue.edu (Chris McCoy) (11/07/90)
I thought I'd post this to the netter's first before I fight with Frame Technology and IBM to see if anyone has experienced similar problems. My configuration: RISC/6000 model 320 32 MB main memory 320 MB disk Gray Scale Graphics Adapter (4-bits) 8508 monitor 3001 update applied I've been experiencing problems with the X server dying and locking my console. Typically, this happens when communicating with remote hosts, but recently when I installed FrameMaker 2.0, it happens quite often. When I run FrameMaker and try to perform any operation on a document or its associated window, the X server dies and I'm hosed. It runs fine on our Xstation, SUN SPARCstation, and Zenith (Frame binary on the IBM). Has anyone else experienced this type of problem? If anyone has FrameMaker running, I'd *really* be interested in hearing from you. I'm convinced the problem lies with the X server on the IBM, but I don't have enough fuel on this problem. -- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Chris McCoy : INTERNET: mccoy@acn.purdue.edu Systems Programmer : UUCP: ...!ecn-ee!mccoy Ag. Computer Network, Purdue Univ. : VOICE: (317) 494-8333 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
scott@prism.gatech.EDU (Scott Holt) (11/07/90)
In article <5973@mace.cc.purdue.edu> lzm@mace.cc.purdue.edu (Chris McCoy) writes: > >I've been experiencing problems with the X server dying and locking >my console. Typically, this happens when communicating with remote >hosts, but recently when I installed FrameMaker 2.0, it happens >quite often. When I run FrameMaker and try to perform any operation >on a document or its associated window, the X server dies and I'm >hosed. It runs fine on our Xstation, SUN SPARCstation, and Zenith >(Frame binary on the IBM). > >Has anyone else experienced this type of problem? If anyone has FrameMaker >running, I'd *really* be interested in hearing from you. I'm convinced >the problem lies with the X server on the IBM, but I don't have enough >fuel on this problem. Well, there are ways to get fuel. First - when your server hangs, find some other route into the system (rlogin, tty) try to force the X server to dump core. Certain signals will do this. You may also find that sending a HUP to the server will cause it to die gracefully - as though you had shut it down. You may also want to force FrameMaker to give a core dump and send that off to Frame. Second, you may find certain xmonitor programs useful. Generate a trace of all the events up to the failure, print it off and let IBM and Frame chew on it. I managed to get xmon to do a nice job of this. I have had similar problems with the X server running on an 8-bit color system. I have, however, been able to run Frame 2.1x as a client (from a remote sparc) with no problem. My problem happened when using Oracle's CASE*Designer as a remote client. I am fairly certain that its the X server, not CASE*Designer, that is the problem - the application runs fine on at least 3 other servers we have tried. I have packaged up an event trace and a core dump of the server to send off to IBM. - Scott -- This is my signature. There are many like it, but this one is mine. Scott Holt Internet: scott@prism.gatech.edu Georgia Tech UUCP: ..!gatech!prism!scott Office of Information Technology, Technical Services
aj4640@dean1.usma.edu (Robertson John MAJ) (11/07/90)
I've had similar problems albeit on a different platform. My X server (the IBM lpp) runs on a PS/2 70 with AIX 1.2. The machine has 16 MB or ram. CLients die for no apparent reason and without warning. FOr example, after running xinit, merely touching either mouse button on the title bar of the term window will cause both it and the window manager to disappear. Only an Alt-AQction sequence brings the the machine back, but now to the UNIX prompt. The problem has been reported but no solutions are on hand yet. Jack
osvaldo@sterope.la.locus.com (Osvaldo Gold) (11/09/90)
In article <5980@mace.cc.purdue.edu> lzm@mace.cc.purdue.edu (Chris McCoy) writes: >... However, the hft is still "stuck" in >graphics mode [when the server dies]. > When your server dies and your console just doesn't want to talk to you, try logging in from another terminal and issue a: tput -Thft clear > /dev/console or tput -Thft clear > /dev/hft This should clear your console screen and put you back in text mode. I do it on the PS2's all the time. -- osvaldo ---------- Osvaldo Gold - Locus Computing Corporation osvaldo@locus.com uunet!lcc!osvaldo DISCLAIMER: Opinions are only my own.
marc@ibmpa.awdpa.ibm.com (Marc Pawliger) (11/10/90)
When X dies, yet the screen still displays what the server looked like before it headed south, and you can't 'hot-key' to another HFT, don't despair. There is a 30 second timeout that will kick in and revert the HFT that hung back to text mode. I realize it's not the best solution, but it beats having to reboot and wait much longer than 30 seconds. These are my opinions, not IBM's, etc etc etc +---Marc Pawliger---IBM Advanced Workstations Division---Palo Alto, CA---+ | Internet: marc@ibminet.awdpa.ibm.com VNET: MARCP at AUSVM6 | | UUCP: uunet!ibmsupt!ibmpa!marc Phone: (415) 855-3493 | +------IBMinet: marc@ibmpa.awdpa.ibm.com------IBM T/L: 465-3493--------+
lzm@mace.cc.purdue.edu (Chris McCoy) (11/11/90)
In <1990Nov9.103301@ibmpa.awdpa.ibm.com> marc@ibmpa.awdpa.ibm.com (Marc Pawliger) writes: >When X dies, yet the screen still displays what the server looked like >before it headed south, and you can't 'hot-key' to another HFT, don't >despair. There is a 30 second timeout that will kick in and revert the >HFT that hung back to text mode. I realize it's not the best solution, >but it beats having to reboot and wait much longer than 30 seconds. What version of AIX are you running? And what platform? Sorry, but this doesn't work on my system. The other day my console locked up and I left it "locked" overnight. The next morning (12 hours later) the screen was still locked. What process monitors the HFT for this 30 second timeout? Tell me more, you've got me curious. -- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Chris McCoy : INTERNET: mccoy@acn.purdue.edu UNIX Systems Programmer : UUCP: ...!ecn-ee!mccoy Ag. Computer Network, Purdue Univ. : VOICE: (317) 494-8333 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=