Sighhhhhhhh.....
As I noted in my last summary, I was uncertain that simply making the
error messages about DDR disappear had really solved my problem. Alas,
it took many more hours of combing through the jumbo patch to find the
real root of my problem.
For those of you who have yet to deal with the jumbo patch
(duv40b00004-19970808), let me give you fair warning - be very careful
which patches you select. If you are using the the dupatch utility, I
would not trust it to tell you what patches are safe for your system.
Fortunately, my 4100 is a test system specifically for DU4.0B and Oracle
7.3.3.3.3. So re-installing was only a headache for me.
It will totally depend on what software subsets you choose for your
original install, how the jumbo patch behaves. Until I really combed
the patch documentation, I experienced problems with ddr_config, kernel
builds, getting CDE to start on the console, etc. There is an option
from dupatch to load ALL patches - I wouldn't if I were you.
The only other observation I can make is that adding the patches in
small groups seemed much more successful.
Good luck and beware large patches,
Susan
> ----------
> From: Susan Rodriguez
> Sent: Tuesday, September 23, 1997 1:33 PM
> To: OSF Managers
> Subject: [Summary] - Invalid database header
>
> Original Question:
> > ----------
> > From: Susan Rodriguez
> > Sent: Monday, September 22, 1997 2:44 PM
> > To: 'OSF Managers'
> > Subject: [Q] - Invalid database header
> >
> > Managers,
> >
> > I have DU4.0b newly installed on my Alpha 4100 5/400 (yes - jumbo
> > patch #4 is on). The following error message appears during system
> > startup (captured from "uerf -R full"):
> >
> >
> > DDR: Invalid Database Header
> > DDR: Unable to access Device_information Database for GLOBALS
> >
> >
> > I believe this is referring to the /etc/ddr.dbase file which is
> > supposed to be a newer and better cam_data.c file.
> >
> > Does anyone know what this error message means?
> >
> > TIA,
> >
> > Susan Rodriguez
> > susrod_at_hbsi.com
> ________
>
> RESOLUTION:
> --------------
>
> Apparently, there is a patch in Jumbo patch #4 that "fixes" some
> problems with ddr_config. ddr_config is a relatively new utility for
> configuring the ddr.db which replaces the old cam_data.c file. The
> patch addresses two problems with ddr_config, the first having to do
> with incomplete entries for some SCSI devices, and the second having
> to
> do with handling the old cam_data.c file. The jumbo patch
> documentation
> details the problem
>
> I re-installed my system after posting this messages. I did not get
> any
> DDR errors until AFTER I put on the jumbo patch. So I backed out of
> just the do_config patch and the errors went away. My best guess
> right
> now is that the compiled /bin/do_config that the patch puts on is not
> compiling the ddr.db file correctly.
>
> Removal of the do_config patch makes the error messages go away. I
> wish
> I was certain that lack of error messages meant no problem exists. HA!
>
> All of my SCSI devices are showing up correctly from the firmware and
> from the OS, and I can mount my RRD45 and RZ29B's to local filesystem
> mounts. Those who responded thought that kernel rebuid would solve
> the
> problem. However, this seems to BE my only real problem - can't get a
> new kernel to build. I'll be posting that one separately (call is in
> to
> DEC) - keeping in mind that they may be related.
>
> Thanks to quick responses from:
>
> Randy Hayman
> Matthias Reichling
>
> Pluggin' away,
>
> susrod_at_hbsi.com
>
Received on Sat Sep 27 1997 - 00:21:52 NZST