[opensuse] Questions about some changes I have noticed with v10.3
I've just installed v10.3 on my Thinkpad and I'm curious about some of the changes that I have noticed. The DVD/CDRW is now /dev/sr0 instead of /dev/hdc. I remember that we used that back when we had the ide-scsi kludge. Is there a reason for changing this? My IDE controllers are now labeled SCSI 0 and SCSI 1 in KInfoCenter My partitions are labeled: /dev/disk/by-id/scsi-SATA_IC25N040ATCS04-CSH405DCJ9x54B-part1 Are all IDE devices now being regarded as SATA/SCSI? I've always felt that it was a bad decision for SATA drives to be labeled sdx. As an example, my Dell has a 9GB SCSI drive in it. If I add a SATA PCI card that would be on a lower slot, my boot drive would be changed from sda to sdb, which would cause a kernel panic(has happened to me before) and crash the system. Other than the problem with my power management on my thinkpad, v10.3 looks pretty good. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 2008/04/08 21:24 (GMT-0400) Larry Stotler apparently typed:
The DVD/CDRW is now /dev/sr0 instead of /dev/hdc. I remember that we used that back when we had the ide-scsi kludge. Is there a reason for changing this? ... Are all IDE devices now being regarded as SATA/SCSI?
Reading release notes before installing is a generally good idea. http://www.suse.com/relnotes/i386/openSUSE/10.3/RELEASE-NOTES.en.html#09 -- "Either the constitution controls the judges, or the judges rewrite the constitution." Judge Robert Bork Team OS/2 ** Reg. Linux User #211409 Felix Miata *** http://fm.no-ip.com/ -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Felix Miata wrote:
On 2008/04/08 21:24 (GMT-0400) Larry Stotler apparently typed:
The DVD/CDRW is now /dev/sr0 instead of /dev/hdc. I remember that we used that back when we had the ide-scsi kludge. Is there a reason for changing this?
...
Note his unanswered question: "REASON for changing"
Are all IDE devices now being regarded as SATA/SCSI?
Reading release notes before installing is a generally good idea. http://www.suse.com/relnotes/i386/openSUSE/10.3/RELEASE-NOTES.en.html#09
I assure you, I did, and he probably did, BUT.... Your answer patently ignores his FIRST question and so does the release notes. They say what it does, not the answer to 'Is there a reason for changing this?' And no mention of how this will affect EXISTING IDE devices with valid OS installations....be they Windoze or as in my case, earlier (then current) versions of SuSE. Changing IDE devices to be treated the same as SATA / SCSI for naming purposes causedme and a whole bunch of others (with mixed IDE/SATA systems)many problems judging from the buglist reports that ensued, not the least of which was during the install of 10.3, it properly installed to my 1st SATA drive (which was bought for the purpose) and put the system onto that drive (then called /dev/sdaN) and the installation on /dev/hdaN was properly ignored at that point. So far, fine, BUT, during the installation procedure, suddenly my /dev/hdaN became /dev/sdaN and the new install was relegated to /dev/sdbN and THEN GRUB was written to point to /dev/sdaN (which used to be my IDE drive) and further installation proceeded to OVERWRITE a perfectly good 10.2 installation which was on the IDE drive and the SATA drive never was completely configured because everything was set for /dev/sdaN and not the renamed /dev/sdbN which contained the first part of the installation. So, I ended up with a hosed 10.2 installation because parts were overwritten, and a hosed 10.3 installation because it was incomplete and GRUB couldn't find a bootable system on either. So, his question seems valid, What was the reason and WHY didn't the readme explain that an IDE drive would be added as the LOWEST /dev/sdX device and the real SATA drives would all be re 'numbered', messing up who knows how many programs that were looking for a specific /dev/sdX drive. No mention that before installation you should disconnect IDE devices to protect them, no mention of changing so that disk drive partitions are accessed by volume name or drive ID name or whatever, but not by /dev/hdX BEFORE attempting installation (and I am not sure that would have prevented the fiasco) but simply saying "Oh, by the way, libata will do such and such". So what? What does this mean to someone trying to install 10.3 on a system that has an IDE system (even Windoze) with an OS on it and because SATA is almost all you can buy now days, or if you already have 4 IDE devices (couple of CDRoms and a master/slave HD) that when you expand your system, you usually will be adding SATA devices. It meant nothing to me, especially as this was a 10.2 to 10.3 'minor' release increment and I wasn't expecting it to hose my good install. None since 7.x through 10.2 had done so to me before and I didn't expect it then. So again, his question was quite valid and quite unanswered by either the release notes, or by your response to him. Richard -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 2008/04/09 07:27 (GMT-0400) Richard Creighton apparently typed:
So again, his question was quite valid and quite unanswered by either the release notes, or by your response to him.
I didn't bother to answer the why because the answer to why isn't useful. The reason is it was decreed it would be thus by kernel developer fiat - libata was decreed by kernel gods to be the replacement for legacy IDE drivers. Since libata uses SCSI as its foundation, libata inherits the most part of SCSI's genes, including its device names and its partitions per disk limit. The release notes do however explain that the SCSI limitations needn't apply to IDE users, but only by properly initiating 10.3 installation to prevent use of libata. -- "Either the constitution controls the judges, or the judges rewrite the constitution." Judge Robert Bork Team OS/2 ** Reg. Linux User #211409 Felix Miata *** http://fm.no-ip.com/ -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Richard Creighton wrote:
Felix Miata wrote:
On 2008/04/08 21:24 (GMT-0400) Larry Stotler apparently typed:
The DVD/CDRW is now /dev/sr0 instead of /dev/hdc. I remember that we used that back when we had the ide-scsi kludge. Is there a reason for changing this?
...
Note his unanswered question: "REASON for changing"
So???
Are all IDE devices now being regarded as SATA/SCSI?
Reading release notes before installing is a generally good idea. http://www.suse.com/relnotes/i386/openSUSE/10.3/RELEASE-NOTES.en.html#09
I assure you, I did, and he probably did, BUT....
Your answer patently ignores his FIRST question and so does the release notes. They say what it does, not the answer to 'Is there a reason for changing this?' And no mention of how this will affect EXISTING IDE devices with valid OS installations....be they Windoze or as in my case, earlier (then current) versions of SuSE.
Changing IDE devices to be treated the same as SATA / SCSI for naming purposes causedme and a whole bunch of others (with mixed IDE/SATA systems)many problems judging from the buglist reports that ensued, not the least of which was during the install of 10.3, it properly installed to my 1st SATA drive (which was bought for the purpose) and put the system onto that drive (then called /dev/sdaN) and the installation on /dev/hdaN was properly ignored at that point. So far, fine, BUT, during the installation procedure, suddenly my /dev/hdaN became /dev/sdaN and the new install was relegated to /dev/sdbN and THEN GRUB was written to point to /dev/sdaN (which used to be my IDE drive) and further installation proceeded to OVERWRITE a perfectly good 10.2 installation which was on the IDE drive and the SATA drive never was completely configured because everything was set for /dev/sdaN and not the renamed /dev/sdbN which contained the first part of the installation. So, I ended up with a hosed 10.2 installation because parts were overwritten, and a hosed 10.3 installation because it was incomplete and GRUB couldn't find a bootable system on either. So, his question seems valid, What was the reason and WHY didn't the readme explain that an IDE drive would be added as the LOWEST /dev/sdX device and the real SATA drives would all be re 'numbered', messing up who knows how many programs that were looking for a specific /dev/sdX drive. No mention that before installation you should disconnect IDE devices to protect them, no mention of changing so that disk drive partitions are accessed by volume name or drive ID name or whatever, but not by /dev/hdX BEFORE attempting installation (and I am not sure that would have prevented the fiasco) but simply saying "Oh, by the way, libata will do such and such". So what? What does this mean to someone trying to install 10.3 on a system that has an IDE system (even Windoze) with an OS on it and because SATA is almost all you can buy now days, or if you already have 4 IDE devices (couple of CDRoms and a master/slave HD) that when you expand your system, you usually will be adding SATA devices. It meant nothing to me, especially as this was a 10.2 to 10.3 'minor' release increment and I wasn't expecting it to hose my good install. None since 7.x through 10.2 had done so to me before and I didn't expect it then. So again, his question was quite valid and quite unanswered by either the release notes, or by your response to him.
If someone explained "WHY" IDE devices are now being treated like SCSI by default, would that change the operation of libata in the slightest? At this point, "Why" the change was made is not important, all that's important is knowing that is was made, and if you need the old IDE functionality, how to get it back. Everything else beyond that is mere trivial history. The kernel devs made a decision, and that *IS* the decision. Arguing about why or why not now is rather pointless -- and I'm a person who subscribes to the view that needlessly limiting the number of partitions on IDE disks...especially with their current sizes....was ill-considered. My personal view is that even if the IDE code was broken, the better solution would have been to fix it. But you know what? They didn't ask me, and as I said, the kernel is what they released not what you, me, or the original poster WISH was released. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Wed, Apr 9, 2008 at 4:27 AM, Richard Creighton
Changing IDE devices to be treated the same as SATA / SCSI for naming purposes causedme and a whole bunch of others (with mixed IDE/SATA systems)many problems judging from the buglist reports that ensued, not the least of which was during the install of 10.3, it properly installed to my 1st SATA drive (which was bought for the purpose) and put the system onto that drive (then called /dev/sdaN) and the installation on /dev/hdaN was properly ignored at that point. So far, fine, BUT, during the installation procedure, suddenly my /dev/hdaN became /dev/sdaN and the new install was relegated to /dev/sdbN and THEN GRUB was written to point to /dev/sdaN (which used to be my IDE drive) and further installation proceeded to OVERWRITE a perfectly good 10.2 installation which was on the IDE drive and the SATA drive never was completely configured because everything was set for /dev/sdaN and not the renamed /dev/sdbN which contained the first part of the installation. So, I ended up with a hosed 10.2 installation because parts were overwritten, and a hosed 10.3 installation because it was incomplete and GRUB couldn't find a bootable system on either. So, his question seems valid, What was the reason and WHY didn't the readme explain that an IDE drive would be added as the LOWEST /dev/sdX device and the real SATA drives would all be re 'numbered', messing up who knows how many programs that were looking for a specific /dev/sdX drive. No mention that before installation you should disconnect IDE devices to protect them, no mention of changing so that disk drive partitions are accessed by volume name or drive ID name or whatever, but not by /dev/hdX BEFORE attempting installation (and I am not sure that would have prevented the fiasco) but simply saying "Oh, by the way, libata will do such and such". So what? What does this mean to someone trying to install 10.3 on a system that has an IDE system (even Windoze) with an OS on it and because SATA is almost all you can buy now days, or if you already have 4 IDE devices (couple of CDRoms and a master/slave HD) that when you expand your system, you usually will be adding SATA devices. It meant nothing to me, especially as this was a 10.2 to 10.3 'minor' release increment and I wasn't expecting it to hose my good install. None since 7.x through 10.2 had done so to me before and I didn't expect it then.
Richard
Paragraphs are your Friends, Richard ;-) -- ----------JSA--------- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
participants (5)
-
Felix Miata -
John Andersen -
Larry Stotler -
Richard Creighton -
Sam Clemens