Mailinglist Archive: opensuse-bugs (13466 mails)
| < Previous | Next > |
[Bug 304657] Yast2 repair doesn' t check or repair system installed on MD Raid
- From: bugzilla_noreply@xxxxxxxxxx
- Date: Wed, 17 Oct 2007 15:03:08 -0600 (MDT)
- Message-id: <20071017210308.C6EF2CC7AD@xxxxxxxxxxxxxxxxxxxxxx>
https://bugzilla.novell.com/show_bug.cgi?id=304657#c29
Richard Creighton <rccj@xxxxxxxxxxx> changed:
What |Removed |Added
----------------------------------------------------------------------------
Attachment #159973|0 |1
is obsolete| |
Attachment #162838|0 |1
is obsolete| |
Attachment #164079|0 |1
is obsolete| |
--- Comment #29 from Richard Creighton <rccj@xxxxxxxxxxx> 2007-10-17 15:03:08
MST ---
Created an attachment (id=179119)
--> (https://bugzilla.novell.com/attachment.cgi?id=179119)
Current state of Yast repair module
This is from an otherwise fully functional GM 10.3 installation of SuSE. It
is using the latest *released* kernel 2.6.22.9-0.4-bigsmp and finally has both
the MD raid and the hardware raid controller from RocketRaid rr174x.ko fully
functional. This gives the machine 2.5TB of pure raid. Due to the IDE/SATA
bug, the IDE drive is still off-line, so the troubleshooting efforts are purely
directed at getting the repair program to recognize and repair the system as
installed on MD raid or other LVM type installations which according to my
E-Mails, are becoming quite common with the larger drives and raid controllers
built-in to the MBds. I feel the repair program is a first line of defense for
the non-technical user and it needs to work.
As you can see from the enclosures, the only thing that does work is that the
repair program detects the partitions on the drives and their types and it
recognizes the fact that there are MD raids in use. It then proceeds to try
to fsck the raw devices rather than the file systems they contain. I believe
this is an error in logic. FSCK should be checking the LOGICAL devices
detected, and only checking the physical device when the logical and physical
are one and the same, ie, don't check devices that have LVM's or MD or other
non raw type file structures. You wouldn't check and try to repair NTFS. You
might check it in some way but you aren't in the business of trying to repair
the drive it lives on unless the user says to reformat the device as a part of
the repair, and then that is another issue entirely.
Let's get back to work on this, OK?
Richard
--
Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
Richard Creighton <rccj@xxxxxxxxxxx> changed:
What |Removed |Added
----------------------------------------------------------------------------
Attachment #159973|0 |1
is obsolete| |
Attachment #162838|0 |1
is obsolete| |
Attachment #164079|0 |1
is obsolete| |
--- Comment #29 from Richard Creighton <rccj@xxxxxxxxxxx> 2007-10-17 15:03:08
MST ---
Created an attachment (id=179119)
--> (https://bugzilla.novell.com/attachment.cgi?id=179119)
Current state of Yast repair module
This is from an otherwise fully functional GM 10.3 installation of SuSE. It
is using the latest *released* kernel 2.6.22.9-0.4-bigsmp and finally has both
the MD raid and the hardware raid controller from RocketRaid rr174x.ko fully
functional. This gives the machine 2.5TB of pure raid. Due to the IDE/SATA
bug, the IDE drive is still off-line, so the troubleshooting efforts are purely
directed at getting the repair program to recognize and repair the system as
installed on MD raid or other LVM type installations which according to my
E-Mails, are becoming quite common with the larger drives and raid controllers
built-in to the MBds. I feel the repair program is a first line of defense for
the non-technical user and it needs to work.
As you can see from the enclosures, the only thing that does work is that the
repair program detects the partitions on the drives and their types and it
recognizes the fact that there are MD raids in use. It then proceeds to try
to fsck the raw devices rather than the file systems they contain. I believe
this is an error in logic. FSCK should be checking the LOGICAL devices
detected, and only checking the physical device when the logical and physical
are one and the same, ie, don't check devices that have LVM's or MD or other
non raw type file structures. You wouldn't check and try to repair NTFS. You
might check it in some way but you aren't in the business of trying to repair
the drive it lives on unless the user says to reformat the device as a part of
the repair, and then that is another issue entirely.
Let's get back to work on this, OK?
Richard
--
Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
| < Previous | Next > |