> Judy Guenther typed (on Wed, Jun 25, 2003 at 09:15:14AM -0700):
> | Bela Lubkin <belalsco.com> wrote in message news:<20030625055830.GI22421sco.com>...
> | > Judy Guenther wrote:
> | >
> | > > Compaq ML570 Server w/4GB Ram, 6-32GB SCSI Hard Drives, Smart-Array
> | > > 5304/128 Controller
> | > >
> | > > During inital install SCO installs correctly, but during first boot
> | > > receives the following: PANIC: srmountfun - error 6 mounting rootdev
> | > > hd(1/42). This occurs each time regardless of which controller the
> | > > hard drives are attached. What has been determined is that after
> | > > system install and during inital boot of O/S if the defbootstr is
> | > > entered at the boot prompt and you quit and don't actually load the
> | > > btld then the system boots correctly and does so from then on. I know
> | > > that the kernel is not re-generated at this time, but is there a
> | > > timing issue or possibly something wrong with my hardware?
> | >
> | > Can you please clarify the above statement? You're apparently using
> | > some boot strings, maybe during ISL, maybe during post-ISL reboots.
> | > Show us the actual strings used at each stage, and clearly state what
> | > happens with each.
> | >
> | > Parts I'm confused about
: "regardless of which controller the hard
> | > drives are attached" -- you only mentioned one controller. The machine
> | > probably has another built in, but don't expect the reader to know that,
> | > spell it out.
> | >
> | > "if the defbootstr is entered at the boot prompt and you quit and don't
> | > actually load the btld" -- haven't a clue what that means.
> | >
> | > If you're using a BTLD (as it sounds like you must be), please name it.
> | >
> | > What are the last few lines of output before the panic? I think
> | > typically there would be some other explanatory error messages before
> | > the "srmountfun" panic (something like "no root disk controller" -- but
> | > please type in the exact text of the last few lines of output).
> | >
> | > >Bela<
> | Sorry for the confusion. We are using EFS 5.58B.
> | Here is the exact error:
> | Error: PANIC: srmountfun - Error 6
> | mounting rootdev hd (1/42)
> | Error 6 opening dumpdev hd (1/41) Dump not completed
> | Warning: ciss <slot 5>: Event notifier disabled.
> | The boot string that was entered when the drives were connected to the
> | 5304 is:
> | defbootstr link=ciss hd=Sdsk Sdsk=ciss(0,0,0,0) Srom=wd(0,0,0)
> | bootstring entered when connected to internal controller was:
> | defbootstr link=cha hd=Sdsk Sdsk=cha(0,0,0,0) Srom=wd(0,0,0)
> | Here is what is happening. The machine is currently up and running.
> | We used the first boot string when loading the OS. On first boot
> | after insallation, we enter the same boot string again.
> Why??? If you link in a particular BTLD at installation, the kernel that
> is compiled when installtion finished *contains that driver*. You no
> longer have to specify its use with a defbootstr command, it's in the
> /stand/unix kernel.
> | At the prompt to enter efs diskette, we quit out and the OS came up.
> | Then, we were unable to see the CDROM drive. Rebooted and did not use
> | the boot string and OS came up and we were able to see CDROM.