Hello,
I have a doubt regarding the "rootdisk" keyword used in jumpstart.
We use the keyword "rootdisk" at two places.
In the "rules" file, we make use of "probe rootdisk" and this is supposed
to fetch the rootdisk and set SI_ROOTDISK environment variable.
Also in "profile" file, we use the keyword "rootdisk".
In the document
http://docs.sun.com/app/docs/doc/819-5778/6n7rbgtan?a=view#preparecustom-25766
it says:
Table 11 - How JumpStart Determines a System's Root Disk (Initial
Installation)
3) If rootdisk is not set and a filesys cwtxdysz size / entry
is specified in the profile, the JumpStart program sets rootdisk
to the disk that is specified in the entry.
4) If rootdisk is not set and a rootdisk.sn entry is
specified in the profile, the JumpStart program searches the system's
disks in kernel probe order for an existing root file system on the
specified slice. If a disk is found, the JumpStart program sets
rootdisk to the found disk.
HOWEVER: In the same document
(http://docs.sun.com/app/docs/doc/819-5778/6n7rbgtan?a=view#preparecustom-25766)
in the section "Custom Jumpstart Environment Variables" section is has the
following:-
" SI_ROOTDISK: The device name of the disk that is represented by the
logical name rootdisk.
The SI_ROOTDISK variable is set when the disksize or the installed
keyword is set to rootdisk in the rules file."
Also in the "Probe Keywords and Value" Section it has the following:
" rootdisk : Returns the name and size in Mbytes of a system's root disk
and sets SI_ROOTDISK. "
Both of these seem to indicate the SI_ROOTDISK env variable would be set when
we use the following entries in the rules file (but it doesn't).
The core problem is that the root disk that is determined by "probe rootdisk"
in the rules files is not the the same as the root disk determined
automatically by the profile.
The pfinstall is searching the system's disks in kernel probe order for an
existing root file system on the specified slice, whereas the rules file is
just looking for the first disk.
The rules file sets SI_ROOTDISK to the first disk, but pfinstall is finding a
root slice on the second disk and so using that and so ignoring the value of
SI_ROOTDISK.
All this results in begin/finish scripts potentially thinking the rootdisk is
different from where the OS was installed.
What needs to happen is that either the function setting SI_ROOTDISK needs to
use the same heuristics as pfinstall, or pfinstall needs to look to, and
honour, the value of SI_ROOTDISK.
So, does it mean, the code is doing something wrong?
Regards,
Narendra
Warm Regards,
S.S.Narendra Kumar
Visit my blogs at:
http://ssnarendrakumar.blogspot.com/
http://blogs.sun.com/roller/page/ssnkumar
___ ___ __ _
/ __/ __ / | //
_\ \ \ \ / /||//
\___/___//_/ |_/
---------------------------------
It's here! Your new message!
Get new email alerts with the free Yahoo! Toolbar.
_______________________________________________
Solaris-Users mailing list
[email protected]
http://node1.filibeto.org/mailman/listinfo/solaris-users
