home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.unix.aix
- Path: sparky!uunet!mcsun!Germany.EU.net!news.netmbx.de!zrz.tu-berlin.de!math.fu-berlin.de!Sirius.dfn.de!chx400!aragorn.unibe.ch!iacrs1!haedener
- From: haedener@iacrs1.unibe.ch (Konrad Haedener)
- Subject: Re: 220 conf'ed for using /usr server
- Message-ID: <1992Aug13.134333.11682@aragorn.unibe.ch>
- Followup-To: <1992Aug13.085828.4829@newssrv.edvz.univie.ac.at>
- Sender: haedener@iacrs1 (Konrad Haedener)
- Reply-To: haedener@iacrs1.unibe.ch
- Organization: University of Berne
- References: <1992Aug13.085828.4829@newssrv.edvz.univie.ac.at>
- Date: Thu, 13 Aug 1992 13:43:33 GMT
- Lines: 205
-
- In article <1992Aug13.085828.4829@newssrv.edvz.univie.ac.at>, tom@bim.itc.univie.ac.at (Tom Kovar) writes:
- |> I'am trying to configure our institute's 220's to use the /usr server, and
- |> have nice problems with it. (Running 3.2)
- |> I am installing the system over network, as we have no tape stations on the
- |> 220's. I used various install images (copied from the original tape, generated
- |> using mksysb of a running 3.2 system, generated using mksysb of a running 3.2
- |> with applied 3.2 fixes, with applied patches upgrading the 3.2 to 3.2.1, and
- |> finaly after applying the patch U401209 and prereq., which should contain
- |> 220 corrections).
- |> After copying a bunch of /usr files, the typical dd-command message comes
- |> (some thousands records in, 0 records out); after that a pax: error message
- |> comes, that the following files weren't found in the archive:
- |> /usr/lib/microcode/8d77.32.52
- |> /usr/lib/microcode/8d77.42.52
- |> /usr/lib/microcode/8d77.44.52
- |> As these files are not on the original installation machine, the only thing
- |> I could do was to require continuation. It tries to generate bos boot file
- |> and boot image on the disk, which aborts with the error message, that it
- |> cannot find the file
- |> /etc//microcode/8d77.32.04
- |> This file *is* on the original install machine, but it is of no help,as the
- |> installation is aborted and there is no boot image on the disk.
- |>
- |> Anyone can help me?
- |> Just a note - according to the manual, I used only the installation image
- |> bos.obj.*; sure it is enuff???
- |>
- |> I'd be immensely gratefull for any kind of help.
- |>
- |> Thanx,
- |> Tom
-
- O.k. People@IBM, take this as a textbook example of how competent your
- AIX support people (over here) are (sorry Mickey C., if you're reading
- this - hand it over to your Swiss colleagues).
-
- I had exactly the same problem and, after loosing a serious amount of time
- and eventually finding a workaround, sent off the following message to IBM:
- =-=-=-=-=-=-
- From haedener Tue Jul 21 09:41:58 1992
- Subject: AIX 3.2.1 Installation Media Defect
- To: aixserv@chvm1.vnet.ibm.com (AIX Support Team)
- Date: Tue, 21 Jul 92 9:41:58 NFT
- Organization: University of Berne, Switzerland
- X-Mailer: ELM [version 2.3 PL11]
-
- MyName: Dr. Konrad Haedener
- MyOrg: University of Berne, Institute for Physical Chemistry
- MyEMailAddr: haedener@iacrs1.unibe.ch
- MyPhone: +41 31 65 42 25
- MyFax: +41 31 65 44 99
- MyMachineS/N: 7011 44-12127
- MyOSlevel: 3.2.1
- MyMedia: 1/4" DC QIC120
- System Number KV207
- Order Number GL93OK
- Serial Number 85AAWQV
- Date 92.06.17 (Denmark)
- Volumes L500275, L500276, L500277, L500278
-
- Severity: 2 (serious loss of time)
-
- Problems: When installing BOS on a mod 220, I get:
-
- pax: ./usr/lib/microcode/8d77.32.52 +
- ./usr/lib/microcode/8d77.42.52 |> not found in archive
- ./usr/lib/microcode/8d77.44.52 +
-
- During bosboot for internal hdisk I get:
-
- /etc/microcode/8d77.32.04 missing
-
- BOS installation remains incomplete and system does not boot
- correctly. When installing BOS on a mod 520 from same media
- and using same tape drive, no errors and no such problems.
-
- When applying selective updates from volume L500278,
- installation stops due to missing prerequisites.
- TOC of L500278 shows:
-
- [ TOC contents deleted ]
-
- Comment: please note that file with sequence number 4 (01:04) is
- missing from TOC (was BOS.U401969 according to media
- description report);
-
- please note that sequence numbers are wrapped to zero after 99,
- i.e.:
-
- (01:98) (01:98)
- (01:99) (01:99)
- (01:00) instead of (01:100)
- (01:01) (01:101)
- (01:02) (01:102)
-
-
- Workaround: - inserting /etc/microcode/8d77.32.04 manually during
- BOS installation on mod 220 (requires another running
- system)
-
- - copying L500278 to disk and manually editing
- /usr/sys/inst.images/.toc; apply missing U401969 from
- diskette.
-
- lscfg(220):
-
- INSTALLED RESOURCE LIST
-
- [ deleted to save bandwidth ]
-
- --
- Konrad Haedener Phone: +41 31 65 42 25
- Institute for Physical Chemistry FAX: +41 31 65 44 99
- University of Berne
- Switzerland haedener@iacrs1.unibe.ch
-
- =-=-=-=-=-=-=-=-=-=-=-=-=
-
- Now, as to the `answer' of IBM:
- =-=-=-=-=-=-=-=-=-=-=-=-=
- Date: Mon, 27 Jul 92 14:51:57 CET
- From: "Martin Tschanz" <MTS@CHVM1.VNET.IBM.COM>
- To: haedener@iacrs1.unibe.ch
- Subject: Tar / Install /Fixes
- Status: RO
-
- Herr Haedener
-
- [ text re another hanging bug/problem deleted ]
-
- Wegen der Installation auf dem Mod 220 habe ich noch unsere Datenbank abge-
- sucht, jedoch keine aehnlichen Probleme gefunden. Unser Herr Muenger, der
- etliche 220 instaliert hat, ist selbst auch noch nicht auf solches gestossen
- (Microcode), wobei er vorausgeschickt hat, dass er keine Installation mit einer
- Gigadisk oder OEM Disk vorgenommen hat.
-
- [ in free translation:
-
- Regarding your installation problem on the mod 220, I have searched our
- database but haven't found any pointers to similar, known problems. My
- colleague Mr Muenger has already installed quite a number of 220s and is
- yet to have problems related to microcode; none of his installations
- was, however, on a gigadisk or an OEM disk.
- ]
-
- [ neither was mine, btw. The `lscfg' output is quite clear about that point ]
-
- Das TOC Problem habe ich mit Herrn Bruderer angeschaut. Wie es aussieht, muss
- in der Routine, die zur Erstellung der 1/4" Tapes gebraucht wird, ein Fehler
- sein. Die verschickten 8mm Tapes aus Daenemark scheinen i.O. zu sein, da wir
- bis jetzt keine aehnliche Reaktion erhalten haben. Ich habe fuer einen Kunden
- letzte Woche 1/4" Tapes auf 8mm umkopiert. Ich werde mich diese Woche bei ihm
- erkundigen, ob er dieselben Probleme hatte.
-
- Ich hoffe, Ihnen geholfen zu haben und werde die Disketten erstellen und zu-
- senden.
-
- Mit freundlichen Gruessen
- Martin Tschanz
-
- [ in free translation:
-
- Together with Mr Bruderer I have had a look into the tape TOC problem.
- As it appears now, there must be a bug in the routine used to copy
- the 1/4" tapes [in Denmark, i.e.]. The 8mm tapes distributed by Denmark
- seem to be o.k., as we have not had similar reactions [from other
- customers]. Last week I have copied 1/4" tapes to 8mm for a customer.
- I will contact him this week to see whether he has had the same problems.
- I hope to have helped you and will send you the diskettes [with quite
- a number of Unnnnn's and IXnnnnn's, i.e.]
- ]
-
- CONCLUSIONS:
-
- o there is a problem with 1/4" install tapes for 3.2.1;
- they contain inconsistent update information, as acknowledged by IBM
- Switzerland. Customer support apparently doesn't care and prefers to
- do statistics, like how many other customers will notice that they are
- not installing the `selective' updates they are thinking they are, but
- rather some others (and certainly loosing a substantial amount of time).
-
- o there is a problem with mod 220s being installed as /usr clients under 3.2.1.
- IBM customer support was not able to reproduce it and thus concludes
- that there is no such problem. Period.
-
- o AIX software distribution is a mess.
-
- o the problem with the IBM AIX support team (CH) is the following:
-
- o customers are a priori wrong;
- o if they can't readily reproduce a problem, there _is_ no problem
- (c.f. statement above); nobody considers solving it as an alternative;
- o they (usually) only work when given precise APAR and PTF numbers
- (learned by customers from sources like comp.unix.aix); they won't
- figure them out by themselves; `working' in this context means
- downloading files and copying them to diskette.
-
- Sorry if I'm stepping on someone's toes, it's by intention: customers
- purchasing equipment from IBM have a right to get adequate support and those reporting bugs shouldn't be treated like fools :-(.
-
- --
- Konrad Haedener Phone: +41 31 65 42 25
- Institute for Physical Chemistry FAX: +41 31 65 44 99
- University of Berne
- Switzerland haedener@iacrs1.unibe.ch
-