gwl@beach.cis.ufl.edu (Gerry Lawrence) (04/26/89)
I've just been (conned) into upgrading to 4.0 on our 3/260. We've got two
Fujitsu 2333's, 8 megs of ram and a 1/4inch tape drive. I can get all the
way through the suninstall forms up until the "Installation" menu item,
which immediately gives:
Bus error: core dumped
Ulch! I've tried different disk partitions, different options, to no
avail. Has anyone had this problem?
By the way we had a disk crash several weeks ago, which left many of the
files trahed, one of the reasons for the upgrade.
[[ Tread very carefully when running suninstall. DO NOT TYPE AHEAD at
all. If you are impatient and hit keys before suninstall is ready to read
them, it may core dump. This was our experience when upgrading a 3/280
with suninstall. This may or may not be the source of your problem, but
it is something to be careful about. --wnl ]]
Thanks in advance
Gerry Lawrence
(normal) gwl@beach.cis.ufl.edu
(stupid) gwl%beach.cis.ufl.edu@bikini.cis.ufl.edu
UUCP: (really stupid) ...!gatech!uflorida!beach.cis.ufl.edu!gwlmlindsey@x102c.harris-atd.com (Lindsey MS 04396) (05/06/89)
gwl@beach.cis.ufl.edu (Gerry Lawrence) writes: >X-Sun-Spots-Digest: Volume 7, Issue 254, message 3 of 12 > >I've just been (conned) into upgrading to 4.0 on our 3/260. We've got two >Fujitsu 2333's, 8 megs of ram and a 1/4inch tape drive. I can get all the >way through the suninstall forms up until the "Installation" menu item, >which immediately gives: > > Bus error: core dumped I ran into a similar problem once installing SUN OS 4.0 on a SUN 3/260 with a vt100 console. It seems that suninstall under 4.0 requires a bit-mapped console for the "Installation" portion of the install. Once the installation was complete, I was able to switch back to a non-SUN console. So if you are using a non-SUN console, try borrowing a SUN monitor from someone and give it a try. Good luck! Steve Lindsey |-) uunet!x102a!mlindsey (407) 727-5893 :-) mlindsey@x102a.harris-atd.com
mcgrew@topaz.rutgers.edu (Charles) (05/06/89)
Hi, gwl@beach.cis.ufl.edu (Gerry Lawrence) writes: > ... I can get all the way through the suninstall forms up > until the "Installation" menu item, which immediately gives: > >Bus error: core dumped > Ulch! I've tried different disk partitions, different > options, to no avail. Has anyone had this problem? Your 'bus error' problem may be due to (believe it or not) a termcap error. We used a v200 for setups on one of our servers, and due to a missing termcap value (for some function that is performed only at install-start), we'd get that result. Our solution was to either use a terminal whose termcap existed (vt100, for instance), or rcp a 'real' termcap file from somewhere else before running suninstall. Hope that helps, Charles
henk@uunet.uu.net (Henk Langeveld) (05/06/89)
gwl@beach.cis.ufl.edu (Gerry Lawrence) writes: >I've just been (conned) into upgrading to 4.0 on our 3/260. We've got two >Fujitsu 2333's, 8 megs of ram and a 1/4inch tape drive. I can get all the >way through the suninstall forms up until the "Installation" menu item, >which immediately gives: > Bus error: core dumped >Ulch! I've tried different disk partitions, different options, to no >avail. Has anyone had this problem? Yes, we had the same trouble in january. I presume you are using ttya as a console ? (i.e. You haven't got a sun console, keyboard and mouse). This is a documented bug, look in the CDB. The problem is that the TERMCAP definition cannot be set from within suninstall, its buffer for the termcap entry is too small. If you set the TERMCAP variable with the correct value before running suninstall, you should have no problems. (This took us, sun NL, two weeks, plus a hardware engineer for half a day at our site swapping boards, until they realised what was happening). >Tread very carefully when running suninstall. DO NOT TYPE AHEAD at >all. If you are impatient and hit keys before suninstall is ready to read >them, it may core dump. This was our experience when upgrading a 3/280 >with suninstall. This may or may not be the source of your problem, but >it is something to be careful about. --wnl I know nothing of problems with typeahead in suninstall, apart from the stupid program reacting a bit slow. [[ Well, it may have just been us. Hard to tell sometimes, and "suninstall" is not something we use that often (so it's not worth the effort of tracking down problems in it). --wnl ]] Henk Langeveld -- Henk Langeveld ( Dept. EUR/EF/AIV; Erasmus University Rotterdam ) henk@{euraiv1,eurtrx}.uucp || +31 10 4081346 langeveld@hroeur5.bitnet ||
todds@uunet.uu.net (Todd Sandor) (05/07/89)
gwl@beach.cis.ufl.edu (Gerry Lawrence) writes: >X-Sun-Spots-Digest: Volume 7, Issue 254, message 3 of 12 >I've just been (conned) into upgrading to 4.0 on our 3/260. We've got two >way through the suninstall forms up until the "Installation" menu item, >which immediately gives: > Bus error: core dumped You don't mension which terminal type your using for the install, but if its a '4.) Other', before running suninstall you must # set noglob # eval `tset -s vt100` # unset noglob # suninstall Then continue on (go slow, as the typeahead feature doesn't work to good), this assumes a vt100, if its not, enter the terminal type in vt100 spot. This is done right after boot the mini-unix (or whatever its called) from the swap partition. Don't forget the post install fixes, the 4.0.1 fixes, their post install fixes, then the itrunc patches, then the security patches, the yapt5.4 fixes, etc....(the story continues....). Its a very good idea to talk to you local Sun support to get them to help you out on which 4.0.1 patches you should install, the README for it describes some patches you may not think you need but really do (ie.winlibs,ldcore), and hopefully they know what they are talking about. We just upgrades our environment (12 Servers, with 80+ workstations), and are till having a few problems that have to be worked out. My recommendation is that unless your forced into it (ie. SMD controllers require it,or for software reasons) don't upgrade to 4.0.1, wait for 4.1. Todd Sandor Voice: (613) 738-1338 ext 2704 P.O. Box 9707 Cognos Incorporated FAX: (613) 738-0002 3755 Riverside Dr. uucp: todds@cognos || uunet!mitel!scs!cognos!todds Ottawa, Ontario Steady as she goes! CANADA K1G 3Z4
perl@step.UUCP (Robert Perlberg) (05/11/89)
gwl@beach.cis.ufl.edu (Gerry Lawrence) writes: > I've just been (conned) into upgrading to 4.0 on our 3/260. We've got two > Fujitsu 2333's, 8 megs of ram and a 1/4inch tape drive. I can get all the > way through the suninstall forms up until the "Installation" menu item, > which immediately gives: > > Bus error: core dumped > > Ulch! I've tried different disk partitions, different options, to no > avail. Has anyone had this problem? We did. Our configuration is a Sun-3/280 with no monitor. We use a vt100-like terminal on ttya as the console. We were told by Sun to type something like this: eval `tset -m:vt100` before running suninstall. Apparently Sun didn't test suninstall enough on such a configuration. Robert Perlberg Dean Witter Reynolds Inc., New York phri!{dasys1 | philabs | mancol}!step!perl -- "I am not a language ... I am a free man!"
jam@philabs.philips.com (John A. Murphy) (05/11/89)
Several people have described the "official" patch from Sun for getting around the ASCII console problem during an install. Another method is to simply do an stty echo (blindly) so you can see what you are typing. Then if you do an ls -F in /usr/etc/install you will see an executable program called "installation". So with a simple # installation your installation will continue just fine. Used this over half a dozen times with no problem. Murf John A. Murphy jam@philabs.philips.com 345 Scarborough Road Briarcliff Manor, NY 10510 One one-trillionith of a surprise: picaboo (914)945-6216
zjat02@uunet.uu.net (Jon A. Tankersley) (05/17/89)
The problem is not that SunOS 4.x expects a Sun monitor, but rather there are some beta-test errors still in the 4.0 FCS tapes (Boy was I pissed when I found them...... Should never have survived Beta testing that one). Before going into suninstall do the following setenv TERM vt100 set noglob eval `tset -s $TERM` unset noglob This will correctly set the environment. Problem also occurs if you are trying to install software on a heterogenous server at a later date than the suninstall. Other problems with 4.0 FCS - top of the head response.... netmask + doesn't have the hostname field in rc.local. rc scripts don't handle setting netmasks on ie1 for gateway systems: rc.single doesn't have hostname set properly on ifconfig for ie1. rc.single expects hostname to be host-gw but that fact isn't commented ANYWHERE I read. host-gw must exist in /etc/hosts. in.named is not run before mount -at nfs. Can cause problems if the hosts requested are not in /etc/hosts but are in the named database. syslog.conf bug with define(LOGHOST, 1) for loghost systems. yp problems with named. (separate patches available). -tank- #include <std/disclaimer.h> /* nobody knows the trouble I .... */ tank@apctrc.trc.amoco.com ..!uunet!apctrc!tank