HP3000-L Archives

August 2000, Week 1

HP3000-L@RAVEN.UTC.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Sletten Kenneth W KPWA <[log in to unmask]>
Reply To:
Sletten Kenneth W KPWA <[log in to unmask]>
Date:
Mon, 7 Aug 2000 18:27:01 -0700
Content-Type:
text/plain
Parts/Attachments:
text/plain (62 lines)
Jeff relates a major war story (makes me wince just to read it):

> Friday arrives.  Spent 17 hours doing the install on one half of
> the mirrored vg00 to leave the 10.20 part intact, ......
> .....  But we goofed somewhere in the split, and the
> 10.20 partition was corrupt.  Ooops.
>
> ......  We tried to boot the 11.00 side of the mirrors but it was
> similarly hosed.  ......
>
> We tried again Sunday for 11.00. .........    Oracle won't run,
> complaining of shared memory. Change kernel parms and
> reboot.  Now it won't run for semaphores.  Tried Apache, won't
> run with shared libraries, static version seems to be OK.   ....

... sounds like great candidate for the "update from Hell" contest..


> Installed > 250Mb of patches from a Custom Patch Manager
> bundle downloaded overnight.  Still no change.

Say, what ??....  That was........   250 **MB**.....  yikes !!...
Our entire MPEXL_SYSTEM_VOLUME_SET is only 1414 MB...

> So, 25 hours x 4 people ......, 7 hours x 3 people ....., 7 more
> hours x 2 people ..... , .....   download of a quarter-gig,  and 4
> more hours x 2 people today still hasn't updated us to 11.00.

let's see.... by my count that's ** 143 hours ** just for system
support people, not counting anything for end-user downtime....

> I want my MPE :-(

When I did an INSTALL of the < long_name > a couple years
ago because PREDICTIVE indicated one of the spindles was
starting to get soft read errors, the complete INSTALL process
took something like 40 minutes (we keep the < long_name >
"HP pure" (all user data on user volumes) )...  O.K...:  That's not
an UPDATE with SUBSYS tape and patches (which as we have
recently seen can lead to its own version of "fun and games")...

But if HP can just make it so that the PATCHIX - STAGEMAN
combination can handle *ALL* patches (including those that
effect NL.PUB.SYS), MPE should (at least in theory) always
have a way to leisurely create a staging area that you can boot
from...   *and* still leave the original BASE group in place "just
in case" you have to go back...  that's a really slick feature (and
would be even slicker if it could handle all update variations).


...  I'm as far removed from advertising as you can get, but even
to me as an amateur it seems like a good marketing department
could come up with some *dynamite* ads comparing MPE and
UNIX in this area;  in MPE's favor....    oops....  I forgot:  HP
doesn't do that....

Maybe Wirt is on the right track after all:
Time for  "Independence Day -- Let the Owl Fly Free"  ??..

;-)  ,
Ken Sletten

ATOM RSS1 RSS2