n_brown@national-institute-for-medical-research.mrc.ac.uk (Nigel Brown) (11/07/90)
Hi.
system: 4D GTX, Irix 3.2
Anyone know what this SYSLOG entry means and how to fix grcond again?
Nov 5 22:37:34 iris2 grcond[2534]: In limbo
Nov 5 22:37:50 iris2 grcond[2534]: Tried and failed 3 times to download graphics subsystem
This happened after our "Computing Dept". wandered in, synced and
reset the iris, then trundled it away to demo something. They do
things like that and now they expect me to fix it...
No idea whether it's relevant, but /dev/gm says:
DEBUG_NOISE at 0x98065E14
Loading PP ucode Version: @(#) PEAPOD 1.2 pp microcode assembler - 6/20/87
Sat Aug 19 18:36:05 1989 user unknown revision(1.123CLOVER2IP4GT)
each time grcond tries to start.
Thanks,
____________________________________________________________________________
SNAIL: Nigel P. Brown, Laboratory of Mathematical Biology,
N.I.M.R., The Ridgeway, Mill Hill, LONDON NW7 1AA, U.K.
Tel: (+44) 081-959 3666 ext 2295
JANET: n_brown@uk.ac.mrc.nimr
INTERNET: n_brown%nimr.mrc.ac.uk@nsfnet-relay.ac.uk
EARN/BITNET: n_brown%nimr.mrc.ac.uk@{cunyvm.cuny.edu|ukacrl}
USENET: ...{uunet|mcvax}!ukc!nimr!n_brown
____________________________________________________________________________doelz@urz.unibas.ch (11/08/90)
In article <14021.9011062201@nimsn41.nimr.mrc.ac.uk>, n_brown@national-institute-for-medical-research.mrc.ac.uk (Nigel Brown) writes: > system: 4D GTX, Irix 3.2 WHICH 4D ? WHICH 3.2 ? > > Anyone know what this SYSLOG entry means and how to fix grcond again? > > Nov 5 22:37:34 iris2 grcond[2534]: In limbo > Nov 5 22:37:50 iris2 grcond[2534]: Tried and failed 3 times to download graphics subsystem It happended to us on a two-processor machine. If this is the bug we encountered, it is fixed in 3.2.2, and 3.3.1. Occurs because the network is on the wrong processor. Try to get 3.3.1, and it will work - at least, it does on our 120. > > This happened after our "Computing Dept". wandered in, synced and > reset the iris, then trundled it away to demo something. They do > things like that and now they expect me to fix it... > At least you know who is guilty :-) Can you restore a backup made before that? > No idea whether it's relevant, but /dev/gm says: > > DEBUG_NOISE at 0x98065E14 > Loading PP ucode Version: @(#) PEAPOD 1.2 pp microcode assembler - 6/20/87 > Sat Aug 19 18:36:05 1989 user unknown revision(1.123CLOVER2IP4GT) > > each time grcond tries to start. > No, this is not relevant. This is an informal message. Contact me for details if you can't get hold of the 3.3.1 release - there is a hack to fix it. Regards Reinhard
n_brown@national-institute-for-medical-research.mrc.ac.uk (Nigel Brown) (11/08/90)
Sorry -- it was a long day... I reported a problem with grcond yesterday, but cocked up the system description. It is a 4D/80 GTX, under Irix 3.2.0. Here's the problem again: > Anyone know what this SYSLOG entry means and how to fix grcond again? > > Nov 5 22:37:34 iris2 grcond[2534]: In limbo > Nov 5 22:37:50 iris2 grcond[2534]: Tried and failed 3 times to download graphics subsystem Thanks, ____________________________________________________________________________ SNAIL: Nigel P. Brown, Laboratory of Mathematical Biology, N.I.M.R., The Ridgeway, Mill Hill, LONDON NW7 1AA, U.K. Tel: (+44) 081-959 3666 ext 2295 JANET: n_brown@uk.ac.mrc.nimr INTERNET: n_brown%nimr.mrc.ac.uk@nsfnet-relay.ac.uk EARN/BITNET: n_brown%nimr.mrc.ac.uk@{cunyvm.cuny.edu|ukacrl} USENET: ...{uunet|mcsun}!ukc!nimr!n_brown ____________________________________________________________________________
n_brown@national-institute-for-medical-research.mrc.ac.uk (Nigel Brown) (11/09/90)
Thanks to all who responded.
The person responsible for the machine going walkabout called in a field
engineer today, who identified the problem as a loose connector inside the
box.
I wasn't present at the operation so can't provide any gory details, except
to say that it is purportedly a common fault.
Moral: Chain your machines to something solid :(
Thanks
nigelmartinm@sgi.com (martin) (11/10/90)
In article <24409.9011080004@nimsn41.nimr.mrc.ac.uk> n_brown@national-institute-for-medical-research.mrc.ac.uk (Nigel Brown) writes: >Sorry -- it was a long day... > >I reported a problem with grcond yesterday, but cocked up the system >description. It is a 4D/80 GTX, under Irix 3.2.0. > >Here's the problem again: >> Anyone know what this SYSLOG entry means and how to fix grcond again? >> >> Nov 5 22:37:34 iris2 grcond[2534]: In limbo >> Nov 5 22:37:50 iris2 grcond[2534]: Tried and failed 3 times to download graphics subsystem > if the machine is a GTX at 3.2 or 3.2.1 the following will be familiar to some: (the following was a post when the problem first surfaced (last december?)) In reference to the problem where a person: - is running 3.2 on an multiprocessor, - is running large jobs in the background, - can't log in to the console with graphics, - can log in NOGRAPHICS and rlogin, - and finds erros like "Tried and failed 3 times to download graphics microcode" in the SYSLOG. There is now a sure-fire temporary solution. It involves reconfiguring the kernel so that the netork runs on processor 0 instead of processor 1. All you need to do is: - edit file /usr/sysgen/master.d/kernel, - change line that reads: int network_processor = 1; to int network_processor = 0; - su to root and cd to /, - type lboot (this will build a new kernel called unix.new in the current directory), - mv unix unix.old, - mv unix.new unix, - sync, init 0, then restart the system. another option is to upgrade to 3.2.2 or 3.2.3 or 3.3.1. -- Martin McDonald Prune Juice - A warrior's drink. SGI - Worf