From: Josh Glover (jmglov@incogen.com)
Date: Fri Aug 23 2002 - 10:50:29 EDT
OK, I have an update on my memory upgrade situation.
As per nearly everyone's suggestion, I tried reseating the new
memory. No dice.
I tried booting up with only 1GB of the new RAM in the system. No
dice.
I tried booting off a CD (excellent suggestion, BTW) to make sure the
hard drive was in no way a factor. Same symptoms, no dice.
So, I finally pulled the new RAM, wrote down the serial number on it,
wrote down the serial number on the old RAM, and called Sun. The
sordid truth that they revealed was beyond our wildest nightmares!
The new RAM (which Sun sent us as a perk for being a good development
partner) is apparently unreliable. The tech whom we talked to said
that mixing the new stuff (serial numbers starting with 5016056) with
the old stuff (serial numbers starting with 5014743) might or might
not work.
We have eight 256MB DIMMs of the 5014743 memory, four in bank A, four
in bank B. The system works fine this way, but when we slap the eight
256MB DIMMs of the 5016056 memory in banks C and D, things become unstable.
So, I ask the list: is anyone using this memory configuration in an
Ultra Enterprise 450 with good results? Please let me know. Advise is,
as always, welcome.
Thanks to everyone for their quick responses:
Kevin Buterbaug
Matt Harris
Leonard, Roger
Robert T Clift
Kevin Metzger
David Foster
Mark Cain
Marc Alvidrez
And my original message:
Quoth Josh Glover (Thu 2002-08-22 03:53:43PM -0400):
> I have an Ultra Enterprise 450 server that we just upgraded from 2GB
> to 4GB of RAM. The installation was no trouble.
>
> However, the machine came back up and tried to boot into Solaris, even
> though I had set auto-boot? to false in the OBP prior to powering the
> machine off for the upgrade. I got a kernel panic with an illegal
> instruction, at which time the machine rebooted itself. This time, it
> stopped at the OK> prompt as I expected. I did a 'boot -r', and
> Solaris started booting. It got as far as displaying the version
> header (SunOS 5.7, if you are curious), then the pipe started
> spinning, but when it gets to the 135/315 degree position, the machine
> locks so hard that a Stop-a does not even work.
>
> After the first such lock, I had to hard cycle the power. When the
> system came back up to the OK> prompt, I checked out the post results
> with .post and looked around in /mc@0,0. Everything looked fine. So I
> tried a 'boot -a -r -s' (interactive, reconfiguration, single-user
> mode), and the kernel started loading fine, then the same thing
> happened: the pipe spun 45 degrees then the machine locked hard.
>
> We installed official Sun RAM, followed the instructions in the
> hardware guide to the T.
>
> Any ideas? I searched the archive and found one similar issue,[1] but
> there was no summary (if he got it fixed and did not summarise, shame
> on him!). I am contacting the OP off-list to see, but I have no great
> hopes. :(
>
>
> TIA and WS,
> Josh
>
>
> [1] http://www.sunmanagers.org/pipermail/sunmanagers/2001-March/001721.html
-- Josh Glover <jmglov@incogen.com> Associate Systems Administrator INCOGEN, Inc. GPG keyID 0x62386967 (Public key available from keyserver.net or pgp.mit.edu) [demime 0.99c.7 removed an attachment of type application/pgp-signature] _______________________________________________ sunmanagers mailing list sunmanagers@sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers
This archive was generated by hypermail 2.1.7 : Wed Apr 09 2008 - 23:24:50 EDT