Sorry I didn't "report" back on this subject sooner.....been VERY busy! Someone else here said that Linux doesn't handle plug and play ESATA nor SATA yet, and that is quite right. It's obviously an area that needs to be addressed SOON as more and more new boxen are being produced with ESATA connections. You have to restart the system with power to the external hard drive and the ESATA port connected for it to work. Once you disconnect the drive, you'll get a "smart" error telling you that the drive is about to fail......annoying but that's about it. Fred -- This message originated from a Linux computer using Open Source software: openSuSE Linux 10.3. No Gates, no Windows....just Linux - STABLE & SECURE! -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Fred A. Miller wrote:
Sorry I didn't "report" back on this subject sooner.....been VERY busy! Someone else here said that Linux doesn't handle plug and play ESATA nor SATA yet, and that is quite right. It's obviously an area that needs to be addressed SOON as more and more new boxen are being produced with ESATA connections.
You have to restart the system with power to the external hard drive and the ESATA port connected for it to work. Once you disconnect the drive, you'll get a "smart" error telling you that the drive is about to fail......annoying but that's about it.
Fred
I had to look up eSATA but that's OK. I learned something new. BUT, you say that Linux doesn't support SATA? My computer boots OpenSuSE 10.3 from a SATA drive. Kubuntu 7.10 boots from hda and I can mount both SATA drives on my system and access directories/files from both. What am I missing? -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 The Thursday 2007-12-06 at 21:00 -0600, Billie Walsh wrote:
BUT, you say that Linux doesn't support SATA?
No. Plug and play, live, sata / esata. Ie, (dis)connect a disk with system running, like a disk on usb. - -- Cheers, Carlos E. R. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.4-svn0 (GNU/Linux) iD8DBQFHWL8ntTMYHG2NR9URAmHfAJ98OeauLxIJB0TTrt41UEoW0kp+JwCfbp7z vbOl0DJJ8GPiQb64/rTON/o= =yVof -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Fri, Dec 07, 2007 at 04:33:42AM +0100, Carlos E. R. wrote:
The Thursday 2007-12-06 at 21:00 -0600, Billie Walsh wrote:
BUT, you say that Linux doesn't support SATA?
No. Plug and play, live, sata / esata. Ie, (dis)connect a disk with system running, like a disk on usb.
Yes, you are correct, however, developers are working on this right now. Luckily, there is not much hardware like this currently out there :) thanks, greg k-h -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Thu, 2007-12-06 at 21:08 -0800, Greg KH wrote:
On Fri, Dec 07, 2007 at 04:33:42AM +0100, Carlos E. R. wrote:
The Thursday 2007-12-06 at 21:00 -0600, Billie Walsh wrote:
BUT, you say that Linux doesn't support SATA?
No. Plug and play, live, sata / esata. Ie, (dis)connect a disk with system running, like a disk on usb.
Yes, you are correct, however, developers are working on this right now. Luckily, there is not much hardware like this currently out there :)
No. You are incorrect on hot swapping SATA. I do this all the time (last checked on a SUSE 10.1 install - soon to be verified on a 10.3 install). 1) unmount the disk 2) remove it. I get this in the system log when I do this: ata1: exception Emask 0x10 SAct 0x0 SErr 0x10000 action 0x42 frozen ata1: (irq_stat 0x00400000, PHY RDY changed) ata1: soft resetting port ata1: SATA link down (SStatus 11 SControl 310) ata1: failed to recover some devices, retrying in 5 secs ata1: hard resetting port ata1: SATA link down (SStatus 0 SControl 310) ata1: failed to recover some devices, retrying in 5 secs ata1: hard resetting port ata1: SATA link down (SStatus 0 SControl 310) ata1.00: disabled ata1: EH complete ata1.00: detaching (SCSI 0:0:0:0) 3) When I insert a new disk, I get this in the system logs: ata1: exception Emask 0x10 SAct 0x0 SErr 0x4040000 action 0x42 frozen ata1: (irq_stat 0x00000040, connection status changed) ata1: waiting for device to spin up (8 secs) ata1: hard resetting port ata1: port is slow to respond, please be patient ata1: port failed to respond (30 secs) ata1: COMRESET failed (device not ready) ata1: hardreset failed, retrying in 5 secs ata1: limiting SATA link speed to 1.5 Gbps ata1: hard resetting port ata1: SATA link up 1.5 Gbps (SStatus 113 SControl 310) ata1.00: ATA-7, max UDMA/133, 625142448 sectors: LBA48 NCQ (depth 31/32) ata1.00: ata1: dev 0 multi count 0 ata1.00: configured for UDMA/133 ata1: EH complete Vendor: ATA Model: Hitachi HDT72503 Rev: V54O Type: Direct-Access ANSI SCSI revision: 05 SCSI device sda: 625142448 512-byte hdwr sectors (320073 MB) sda: Write Protect is off sda: Mode Sense: 00 3a 00 00 SCSI device sda: drive cache: write back SCSI device sda: 625142448 512-byte hdwr sectors (320073 MB) sda: Write Protect is off sda: Mode Sense: 00 3a 00 00 SCSI device sda: drive cache: write back sda: sda1 sd 0:0:0:0: Attached scsi disk sda sd 0:0:0:0: Attached scsi generic sg0 type 0 My system has 4 removable SATA disks. So I have a udev rule to mount them in a predictable way, no matter which of the 4 are really present. And, since I run KDE,I also get a dialog when the new disk is inserted. Be sure to unmount the disks first. Otherwise the mount table retains old information about disks no longer mounted. If this has happened, clean it up with: umount -a mount -a -- Roger Oberholtzer OPQ Systems / Ramböll RST Ramböll Sverige AB Kapellgränd 7 P.O. Box 4205 SE-102 65 Stockholm, Sweden Office: Int +46 8-615 60 20 Mobile: Int +46 70-815 1696 -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
|-----Original Message----- |From: Roger Oberholtzer [mailto:roger@opq.se] |My system has 4 removable SATA disks. So I have a udev rule to |mount them in a predictable way, no matter which of the 4 are |really present. I little question on the side: My problem is If I leave the esata connected drives on during a reboot, they come up as sda,sdb forcing the internals to become sdc,sdd. How can this be avoided, On redhat we earlier used mount labels, but I've not been using it for years. I also wonder if there are some external esata cabinets that can power on and off drives via software, I only use them for backups approx 6hours a week, and they generate lots of heat. -- Morten Bjoernsvik, Oslo, Norway
On Fri, 2007-12-07 at 09:54 +0100, Morten Bjørnsvik wrote:
|-----Original Message----- |From: Roger Oberholtzer [mailto:roger@opq.se] |My system has 4 removable SATA disks. So I have a udev rule to |mount them in a predictable way, no matter which of the 4 are |really present.
I little question on the side:
My problem is If I leave the esata connected drives on during a reboot, they come up as sda,sdb forcing the internals to become sdc,sdd. How can this be avoided, On redhat we earlier used mount labels, but I've not been using it for years.
Hard to tell. I have my SATA controller in AHCI mode (set in the BIOS on my supermicro computer). Then the disks get an address based on the SATA connector they are attached to. This address does not change if a disk is or is not present. Of course, if you play with this BIOS setting, your disks may initially move. So you may need to change the boot info. But that should only happen when you go to AHCI mode. The libata folk imply that they like AHCI mode. I bet (just a guess) they test with that setup. This allows me to use udev rules for my removable disks. I would also think it would keep your boot disks from moving around. My rules for the four removable disks I have are: SUBSYSTEM=="block", BUS=="scsi", KERNEL=="sd*1", ID=="0:0:0:0", SYMLINK="cameraA 0:0:0:0" SUBSYSTEM=="block", BUS=="scsi", KERNEL=="sd*[2-9]", ID=="0:0:0:0", SYMLINK="cameraA_p%n" SUBSYSTEM=="block", BUS=="scsi", KERNEL=="sd*1", ID=="1:0:0:0", SYMLINK="cameraB 1:0:0:0" SUBSYSTEM=="block", BUS=="scsi", KERNEL=="sd*[2-9]", ID=="1:0:0:0", SYMLINK="cameraB_p%n" SUBSYSTEM=="block", BUS=="scsi", KERNEL=="sd*1", ID=="2:0:0:0", SYMLINK="cameraC 2:0:0:0" SUBSYSTEM=="block", BUS=="scsi", KERNEL=="sd*[2-9]", ID=="2:0:0:0", SYMLINK="cameraC_p%n" SUBSYSTEM=="block", BUS=="scsi", KERNEL=="sd*1", ID=="3:0:0:0", SYMLINK="cameraD 3:0:0:0" SUBSYSTEM=="block", BUS=="scsi", KERNEL=="sd*[2-9]", ID=="3:0:0:0", SYMLINK="cameraD_p%n" (Note that SATA disks are part of the 'SCSI' system in Linux, so the BUS is 'scsi'.) What this does is, for each bay, make a sym link that refers to the drive bay. So, no matter which disks are or are not present, if a disk is in a certain bay, it can have a consistent name based on the bay it is in. In my case, names like /dev/cameraA. My naming deals with all partitions on the disk. Then, I use these symbolic names in mount tables and all. I do not care if they are really sdb1 or sdc4. I hope some of this helps.
-- Morten Bjoernsvik, Oslo, Norway
��Nry隊Z)z{.ﮞ˛m)z{.+Z+ib*'jW(fvǦj)hǾi
-- Roger Oberholtzer OPQ Systems / Ramböll RST Ramböll Sverige AB Kapellgränd 7 P.O. Box 4205 SE-102 65 Stockholm, Sweden Office: Int +46 8-615 60 20 Mobile: Int +46 70-815 1696 -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Thanks Roger I'll test out and report back to mailinglist -- MortenB |-----Original Message----- |From: Roger Oberholtzer [mailto:roger@opq.se] |Sent: 7. desember 2007 10:32 |To: opensuse@opensuse.org |Subject: RE: [opensuse] ESATA | |On Fri, 2007-12-07 at 09:54 +0100, Morten Bjørnsvik wrote: |> |-----Original Message----- |> |From: Roger Oberholtzer [mailto:roger@opq.se] My system has 4 |> |removable SATA disks. So I have a udev rule to mount them in a |> |predictable way, no matter which of the 4 are really present. |> |> I little question on the side: |> |> My problem is If I leave the esata connected drives on during a |> reboot, they come up as sda,sdb forcing the internals to become |> sdc,sdd. How can this be avoided, On redhat we earlier used |mount labels, but I've not been using it for years. | |Hard to tell. I have my SATA controller in AHCI mode (set in |the BIOS on my supermicro computer). Then the disks get an |address based on the SATA connector they are attached to. This |address does not change if a disk is or is not present. Of |course, if you play with this BIOS setting, your disks may |initially move. So you may need to change the boot info. |But that should only happen when you go to AHCI mode. The |libata folk imply that they like AHCI mode. I bet (just a |guess) they test with that setup. | |This allows me to use udev rules for my removable disks. I |would also think it would keep your boot disks from moving around. | |My rules for the four removable disks I have are: | |SUBSYSTEM=="block", BUS=="scsi", KERNEL=="sd*1", |ID=="0:0:0:0", SYMLINK="cameraA 0:0:0:0" |SUBSYSTEM=="block", BUS=="scsi", KERNEL=="sd*[2-9]", |ID=="0:0:0:0", SYMLINK="cameraA_p%n" | |SUBSYSTEM=="block", BUS=="scsi", KERNEL=="sd*1", |ID=="1:0:0:0", SYMLINK="cameraB 1:0:0:0" |SUBSYSTEM=="block", BUS=="scsi", KERNEL=="sd*[2-9]", |ID=="1:0:0:0", SYMLINK="cameraB_p%n" | |SUBSYSTEM=="block", BUS=="scsi", KERNEL=="sd*1", |ID=="2:0:0:0", SYMLINK="cameraC 2:0:0:0" |SUBSYSTEM=="block", BUS=="scsi", KERNEL=="sd*[2-9]", |ID=="2:0:0:0", SYMLINK="cameraC_p%n" | |SUBSYSTEM=="block", BUS=="scsi", KERNEL=="sd*1", |ID=="3:0:0:0", SYMLINK="cameraD 3:0:0:0" |SUBSYSTEM=="block", BUS=="scsi", KERNEL=="sd*[2-9]", |ID=="3:0:0:0", SYMLINK="cameraD_p%n" | | |(Note that SATA disks are part of the 'SCSI' system in Linux, |so the BUS is 'scsi'.) What this does is, for each bay, make a |sym link that refers to the drive bay. So, no matter which |disks are or are not present, if a disk is in a certain bay, |it can have a consistent name based on the bay it is in. In my |case, names like /dev/cameraA. My naming deals with all |partitions on the disk. Then, I use these symbolic names in |mount tables and all. I do not care if they are really sdb1 or sdc4. | |I hope some of this helps. | |> |> -- |> Morten Bjoernsvik, Oslo, Norway |> |> ��Nry隊Z)z{.ﮞ˛m)z{.+Z+ib*'jW(fvǦj)hǾi |-- |Roger Oberholtzer | |OPQ Systems / Ramböll RST | |Ramböll Sverige AB |Kapellgränd 7 |P.O. Box 4205 |SE-102 65 Stockholm, Sweden | |Office: Int +46 8-615 60 20 |Mobile: Int +46 70-815 1696 | |-- |To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For |additional commands, e-mail: opensuse+help@opensuse.org | | |
On 2007/12/07 09:54 (GMT+0100) Morten Bjørnsvik apparently typed:
My problem is If I leave the esata connected drives on during a reboot, they come up as sda,sdb forcing the internals to become sdc,sdd. How can this be avoided,
There may be motherboard BIOS setting(s) to do it. It seems if BIOS were smart, they would default to external connectors = last found. Are your external ports on a PCI card? Are your internals actually PATA, being supported by the OS as SCSI via libata?
On redhat we earlier used mount labels, but I've not been using it for years.
AFAIK, all current and recent Linux distros support mount by-label. Fedora 7-up installation enforces mount by-label. IMO by-label is far superior in user friendliness to mounting by-UUID or device ID, keeping fstab comprehensible and more compact. I mount my 10.2 root by-label, but all other partitions by device name. So far I've only used eSATA for full partition backups, and so haven't needed to mount any eSATA partitions. -- " Our Constitution was made only for a moral and religious people. It is wholly inadequate to the government of any other." John Adams Team OS/2 ** Reg. Linux User #211409 Felix Miata *** http://mrmazda.no-ip.com/ -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Fri, 2007-12-07 at 09:27 -0500, Felix Miata wrote:
On 2007/12/07 09:54 (GMT+0100) Morten Bjørnsvik apparently typed:
My problem is If I leave the esata connected drives on during a reboot, they come up as sda,sdb forcing the internals to become sdc,sdd. How can this be avoided,
There may be motherboard BIOS setting(s) to do it. It seems if BIOS were smart, they would default to external connectors = last found. Are your external ports on a PCI card? Are your internals actually PATA, being supported by the OS as SCSI via libata?
I think this is a BIOS thing if it is on the motherboard. Otherwise it is a card BIOS setting. If the SATA controller card has this.
On redhat we earlier used mount labels, but I've not been using it for years.
AFAIK, all current and recent Linux distros support mount by-label. Fedora 7-up installation enforces mount by-label. IMO by-label is far superior in user friendliness to mounting by-UUID or device ID, keeping fstab comprehensible and more compact. I mount my 10.2 root by-label, but all other partitions by device name. So far I've only used eSATA for full partition backups, and so haven't needed to mount any eSATA partitions.
In our use of hot swapped SATA disks, there are literally hundreds of them. Labels are not practical. I also tried the serial number of the disk via udev. That works as well. But it was impractical to rely on this in our odd use. So we opted for sensing the drive bay, which fits our setup. -- Roger Oberholtzer OPQ Systems / Ramböll RST Ramböll Sverige AB Kapellgränd 7 P.O. Box 4205 SE-102 65 Stockholm, Sweden Office: Int +46 8-615 60 20 Mobile: Int +46 70-815 1696 -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 The Friday 2007-12-07 at 09:27 -0500, Felix Miata wrote:
AFAIK, all current and recent Linux distros support mount by-label. Fedora
Even grub in 10.3 supports mounting by label. I tried yesterday: title MAIN openSUSE 10.3 (default-label) root (hd0,5) kernel /vmlinuz root=/dev/disk/by-label/160_root vga=0x317 hwprobe=-modules.pata resume=/dev/hda5 splash=verbose showopts apic initrd /initrd - -- Cheers, Carlos E. R. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.4-svn0 (GNU/Linux) iD8DBQFHWWe5tTMYHG2NR9URAquKAKCOkXATamSX7mFDtz/+ciK2xdcIYgCgk5bW mVuWlK7+B7/mSCYE2442ry0= =Ueci -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 2007/12/07 16:33 (GMT+0100) Carlos E. R. apparently typed:
The Friday 2007-12-07 at 09:27 -0500, Felix Miata wrote:
AFAIK, all current and recent Linux distros support mount by-label
Even grub in 10.3 supports mounting by label. I tried yesterday:
title MAIN openSUSE 10.3 (default-label) root (hd0,5) kernel /vmlinuz root=/dev/disk/by-label/160_root vga=0x317 hwprobe=-modules.pata resume=/dev/hda5 splash=verbose showopts apic
ISTR the YaST installer puts all kernel parameters used during installation on the kernel lines in menu.lst, but does anyone know if hwprobe=-modules.pata has any meaning or relevance to anything other than the YaST installer? -- " Our Constitution was made only for a moral and religious people. It is wholly inadequate to the government of any other." John Adams Team OS/2 ** Reg. Linux User #211409 Felix Miata *** http://mrmazda.no-ip.com/ -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 The Friday 2007-12-07 at 11:26 -0500, Felix Miata wrote:
Even grub in 10.3 supports mounting by label. I tried yesterday:
title MAIN openSUSE 10.3 (default-label) root (hd0,5) kernel /vmlinuz root=/dev/disk/by-label/160_root vga=0x317 hwprobe=-modules.pata resume=/dev/hda5 splash=verbose showopts apic
ISTR the YaST installer puts all kernel parameters used during installation on the kernel lines in menu.lst, but does anyone know if hwprobe=-modules.pata has any meaning or relevance to anything other than the YaST installer?
It dissables using the new libata for "traditional" PATA disks. I require that option as I have more than 15 partitions. - -- Cheers, Carlos E. R. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.4-svn0 (GNU/Linux) iD8DBQFHWYMTtTMYHG2NR9URAslWAJ9WZf/6DVYA1FnZhz2w0nh3l8l1ngCbB1Fh tl3KEKfFy3ZdncMGWAs8o70= =rXwO -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
^^^^^^^^^^^^^^^^^^^^^^ After Installation Completed ********************** On 2007/12/07 12:29 (GMT-0500) Carlos E. R. apparently typed:
The Friday 2007-12-07 at 11:26 -0500, Felix Miata wrote:
other than the YaST installer? ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ It dissables using the new libata for "traditional" PATA disks. I require that option as I have more than 15 partitions.
# cat /etc/SuSE-release openSUSE 10.3 (i586) VERSION = 10.3 # uname -r 2.6.22.5-31-default # cat /proc/cmdline root=/dev/disk/by-label/H16A-22factory noapic noresume splash=0 3 vga=788 # cat /boot/grub/menu.lst default 0 timeout 15 gfxmenu (hd0,21)/boot/message title openSUSE Factory -default kernel root (hd0,21) kernel /boot/vmlinuz root=/dev/disk/by-label/H16A-22factory showopts noapic noresume splash=0 3 vga=788 initrd /boot/initrd # mount /dev/hda22 on / type ext3 (rw,noatime,acl) /dev/hda7 on /disks/hda/boot type ext2 (rw,noatime,acl) /dev/hda14 on /home type ext3 (rw,noatime,acl) /dev/hda17 on /pub type ext3 (rw,noatime,acl) /dev/hda18 on /srv type ext3 (rw,noatime,acl) /dev/hda19 on /usr/local type ext3 (rw,noatime,acl) /dev/hda20 on /usr/src type ext3 (rw,noatime,acl) -- " Our Constitution was made only for a moral and religious people. It is wholly inadequate to the government of any other." John Adams Team OS/2 ** Reg. Linux User #211409 Felix Miata *** http://mrmazda.no-ip.com/ -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 The Friday 2007-12-07 at 14:08 -0500, Felix Miata wrote:
On 2007/12/07 12:29 (GMT-0500) Carlos E. R. apparently typed:
The Friday 2007-12-07 at 11:26 -0500, Felix Miata wrote:
other than the YaST installer? ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Ok, ok, but it was Yast who wrote that parameter.
kernel /boot/vmlinuz root=/dev/disk/by-label/H16A-22factory showopts noapic noresume splash=0 3 vga=788
# mount /dev/hda22 on / type ext3 (rw,noatime,acl) /dev/hda7 on /disks/hda/boot type ext2 (rw,noatime,acl) /dev/hda14 on /home type ext3 (rw,noatime,acl) /dev/hda17 on /pub type ext3 (rw,noatime,acl) /dev/hda18 on /srv type ext3 (rw,noatime,acl) /dev/hda19 on /usr/local type ext3 (rw,noatime,acl) /dev/hda20 on /usr/src type ext3 (rw,noatime,acl)
That is indeed a surprise to me. I'm utterly surprised... I don't understand. I'll have to try myself... Does it say anything in /var/log/boot.msg? I don't know how to detect what library it is using, libata or the classic one, whatever the name is. Perhaps: nimrodel:~ # lsmod | grep -i ide ide_cd 40324 0 cdrom 37020 1 ide_cd ide_disk 20480 28 ide_core 122948 3 ide_cd,ide_disk,piix - -- Cheers, Carlos E. R. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.4-svn0 (GNU/Linux) iD8DBQFHWbL0tTMYHG2NR9URAlHhAJ44Ml/CO/SgfS+kiQmvrMXJ9LhcEgCfdvsi 38LX6lcaRAJWYNHHivQI5sY= =qELX -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 12/08/2007 04:54 AM, Carlos E. R. wrote:
The Friday 2007-12-07 at 14:08 -0500, Felix Miata wrote:
On 2007/12/07 12:29 (GMT-0500) Carlos E. R. apparently typed:
The Friday 2007-12-07 at 11:26 -0500, Felix Miata wrote:
other than the YaST installer? ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Ok, ok, but it was Yast who wrote that parameter.
kernel /boot/vmlinuz root=/dev/disk/by-label/H16A-22factory showopts noapic noresume splash=0 3 vga=788
# mount /dev/hda22 on / type ext3 (rw,noatime,acl) /dev/hda7 on /disks/hda/boot type ext2 (rw,noatime,acl) /dev/hda14 on /home type ext3 (rw,noatime,acl) /dev/hda17 on /pub type ext3 (rw,noatime,acl) /dev/hda18 on /srv type ext3 (rw,noatime,acl) /dev/hda19 on /usr/local type ext3 (rw,noatime,acl) /dev/hda20 on /usr/src type ext3 (rw,noatime,acl)
That is indeed a surprise to me. I'm utterly surprised... I don't understand. I'll have to try myself...
I wonder if hwprobe=-modules.pata is passed to the installer, it includes the older module in the initrd and perhaps blacklists the libata one. Surely the initrd controls what module loads on boot to access the hard drives. -- Joe Morris Registered Linux user 231871 running openSUSE 10.3 x86_64 -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 The Saturday 2007-12-08 at 11:33 +0800, Joe Morris (NTM) wrote:
I wonder if hwprobe=-modules.pata is passed to the installer, it includes the older module in the initrd and perhaps blacklists the libata one. Surely the initrd controls what module loads on boot to access the hard drives.
Let's see... no, it seems both types are included. nimrodel:/usr/src/linux # mkinitrd ... Kernel image: /boot/vmlinuz-2.6.22.12-0.1-cer Initrd image: /boot/initrd-2.6.22.12-0.1-cer Root device: /dev/hdd6 (mounted on / as ext3) Resume device: /dev/hda5 Kernel Modules: processor thermal ide-core piix fan jbd mbcache ext3 edd reiserfs generic scsi_mod libata ata_piix ata_generic ide-disk usbcore ohci-hcd uhci-hcd ehci-hcd ff-memless hid usbhid Features: block usb resume.userspace resume.kernel Bootsplash: SuSE (1024x768) lddlibc4: cannot read header from `/usr/local/sbin/test': No such file or directory 46890 blocks I see it includes libata and also ide-disk, ide-core... but libata is not loaded on memory. Funny... raid modules are not included. I thought they should :-? I have a small raid partition for testing. That could explain an error I saw. - -- Cheers, Carlos E. R. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.4-svn0 (GNU/Linux) iD8DBQFHWnqGtTMYHG2NR9URAlqdAKCDvxSZmfbjFDkhfNwg1Jn2C3gg4ACgl6jC hBvNSiDsyxky8Sq5IxEbfO4= =M4gB -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 12/08/2007 07:05 PM, Carlos E. R. wrote:
nimrodel:/usr/src/linux # mkinitrd ...
Kernel image: /boot/vmlinuz-2.6.22.12-0.1-cer Initrd image: /boot/initrd-2.6.22.12-0.1-cer Root device: /dev/hdd6 (mounted on / as ext3) Resume device: /dev/hda5 Kernel Modules: processor thermal ide-core piix fan jbd mbcache ext3 edd reiserfs generic scsi_mod libata ata_piix ata_generic ide-disk usbcore ohci-hcd uhci-hcd ehci-hcd ff-memless hid usbhid Features: block usb resume.userspace resume.kernel Bootsplash: SuSE (1024x768) lddlibc4: cannot read header from `/usr/local/sbin/test': No such file or directory 46890 blocks
I see it includes libata and also ide-disk, ide-core... but libata is not loaded on memory.
Is it by any chance blacklisted? It appears it is loaded into your initrd because ata_piix and maybe ata_generic are in your initrd line in /etc/sysconfig/kernel file. Are those also not loaded according to lsmod? I wonder if you removed those from your INITRD_MODULES line, and rebuilt your initrd, if those 3 would be gone, and it would boot without the hwprobe line?
Funny... raid modules are not included. I thought they should :-? I have a small raid partition for testing. That could explain an error I saw.
I have raid1 module listed in mine (it better, I have a raid1 / and home). -- Joe Morris Registered Linux user 231871 running openSUSE 10.3 x86_64 -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 The Saturday 2007-12-08 at 19:49 +0800, Joe Morris (NTM) wrote:
I see it includes libata and also ide-disk, ide-core... but libata is not loaded on memory.
Is it by any chance blacklisted?
No, it is not. But you can not check my setup much, because I have the hwprobe=-modules.pata token on grub. We have to ask Felix. And the module (which ever one) is surely loaded very early on boot.
Funny... raid modules are not included. I thought they should :-? I have a small raid partition for testing. That could explain an error I saw. I have raid1 module listed in mine (it better, I have a raid1 / and home).
I don't, but mine is only a data partition I use to learn things about raid. And I had problems to mount it during boot (on 10.2 it did), till I defined the /etc/mdadm.conf, which was previously undefined and it worked. Could be related to the raid1 module on initrd, dunno. - -- Cheers, Carlos E. R. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.4-svn0 (GNU/Linux) iD8DBQFHWoeVtTMYHG2NR9URAjBSAJwNgW6nTo78aqPy5EY01uo+0t7NOwCfYh+0 yVKYY0v/CL71/tMztZwjVqk= =xyPl -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 The Saturday 2007-12-08 at 13:01 +0100, Carlos E. R. wrote:
But you can not check my setup much, because I have the hwprobe=-modules.pata token on grub. We have to ask Felix.
I rebooted without that option, and libata is not loaded, and I can access all my partitions. Funny! - -- Cheers, Carlos E. R. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.4-svn0 (GNU/Linux) iD8DBQFHWpzxtTMYHG2NR9URAipsAJ0bzYu9dY6lcjxrcj+brQgpAi6rhACfT4HU qFju+gAK648XJoirJVpyO4o= =FPXf -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 2007/12/08 08:32 (GMT-0500) Carlos E. R. apparently typed:
The Saturday 2007-12-08 at 13:01 +0100, Carlos E. R. wrote:
hwprobe=-modules.pata
I rebooted without that option, and libata is not loaded, and I can access all my partitions.
Now maybe the thread can get onto answering the reason why I started it, and the thread subject: "does anyone know if hwprobe=-modules.pata has any meaning or relevance to anything other than the YaST installer?" -- " Our Constitution was made only for a moral and religious people. It is wholly inadequate to the government of any other." John Adams Team OS/2 ** Reg. Linux User #211409 Felix Miata *** http://mrmazda.no-ip.com/ -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 12/08/2007 10:10 PM, Felix Miata wrote:
On 2007/12/08 08:32 (GMT-0500) Carlos E. R. apparently typed:
I rebooted without that option, and libata is not loaded, and I can access all my partitions.
Now maybe the thread can get onto answering the reason why I started it, and the thread subject: "does anyone know if hwprobe=-modules.pata has any meaning or relevance to anything other than the YaST installer?"
I sure don't understand your statement Felix. What do you think we were trying to do? It is obvious by what Carlos discovered by trial is that if the libata modules (in his case ata_piix and ata_generic) are not in the initrd, it boots even without the hwprobe line (if I understood what he did). It is also obvious with it in the boot arguments, that even if the libata modules ARE in the initrd, it keeps them from loading and allows the regular ide drivers to control. So hwprobe=-modules.pata does have relevance to more than the installer, but it is possible to work without it. It may not be able for the installer to work without it, since you can not change its initrd. At least that is my understanding AFTER the things Carlos has been willing to test. Mine would not be a good one to use to test, as I use libata. -- Joe Morris Registered Linux user 231871 running openSUSE 10.3 x86_64 -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 2007/12/08 22:46 (GMT+0800) Joe Morris (NTM) apparently typed:
On 12/08/2007 09:10:26 (GMT-0500), Felix Miata wrote:
On 2007/12/08 14:32 (GMT+0100) Carlos E. R. apparently typed:
I rebooted without that option, and libata is not loaded, and I can access all my partitions.
Now maybe the thread can get onto answering the reason why I started it, and the thread subject: "does anyone know if hwprobe=-modules.pata has any meaning or relevance to anything other than the YaST installer?"
I sure don't understand your statement Felix. What do you think we were trying to do? It is obvious by what Carlos discovered by trial is that if the libata modules (in his case ata_piix and ata_generic) are not in the initrd, it boots even without the hwprobe line (if I understood what he did).
NAICT, Carlos "discovered by trial" what I had already demonstrated in http://lists.opensuse.org/opensuse/2007-12/msg00810.html that hwprobe=-modules.pata serves no apparent function as a kernel parameter for booting an already installed system. I was trying to thwart further discussion that seemed focused on determining something already shown.
It is also obvious with it in the boot arguments, that even if the libata modules ARE in the initrd, it keeps them from loading and allows the regular ide drivers to control. So hwprobe=-modules.pata does have relevance to more than the installer,
I haven't yet seen it demonstrated in the thread what post-installation relevance hwprobe=-modules.pata exhibits. I was trying to push discussion to discovering whatever might exist.
but it is possible to work without it. It may not be able for the installer to work without it, since you can not change its initrd. At least that is my understanding AFTER the things Carlos has been willing to test. Mine would not be a good one to use to test, as I use libata.
Have you guys been saying that hwprobe=-modules.pata has some relevance to how an initrd is built? I do have one system running 10.2 and Factory that does have both PATA (NForce2) and (e)SATA (VT6420 "RAID") controllers, though I don't normally have any SATA disks attached to it. http://www.biostar-usa.com/mbdetails.asp?model=M7NCD%20ULTRA If I knew something to try to demonstrate hwprobe=-modules.pata has some post-installation relevance, I could. You might note I simultaneously asked the exact same original question on the Factory list, but that as yet has shown no interest from any devs who might know the answer. -- " Our Constitution was made only for a moral and religious people. It is wholly inadequate to the government of any other." John Adams Team OS/2 ** Reg. Linux User #211409 Felix Miata *** http://mrmazda.no-ip.com/ -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 The Saturday 2007-12-08 at 09:10 -0500, Felix Miata wrote:
Now maybe the thread can get onto answering the reason why I started it, and the thread subject: "does anyone know if hwprobe=-modules.pata has any meaning or relevance to anything other than the YaST installer?"
I have no idea. I only know that I booted with: <5>Kernel command line: root=/dev/disk/by-id/ata-ST3160021A_5JS4VV1F-part6 vga=0x317 resume=/dev/disk/by-label/320_swap splash=verbose apic and that my initrd file contains libata, but nevertheless, my kernel doesn't use libata - which is good for me. And it must do so before exploring the partitions. Somehow it must detect the need, and uses the old method. - -- Cheers, Carlos E. R. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.4-svn0 (GNU/Linux) iD8DBQFHWsSktTMYHG2NR9URAm0hAJ9fhIsT7rto/XQioEdhdyhW4SnolQCffP5s NdqmTJ3P2Za1ckOImUtXksk= =2eQH -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 12/06/2007 Carlos E. R. wrote:
The Thursday 2007-12-06 at 21:00 -0600, Billie Walsh wrote:
BUT, you say that Linux doesn't support SATA?
No. Plug and play, live, sata / esata. Ie, (dis)connect a disk with system running, like a disk on usb.
- -- Cheers, Carlos E. R.
OK. I wasn't sure. Thanks -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Carlos E. R. wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
The Thursday 2007-12-06 at 21:00 -0600, Billie Walsh wrote:
BUT, you say that Linux doesn't support SATA?
No. Plug and play, live, sata / esata. Ie, (dis)connect a disk with
All the SATA is plug-and-play, just like almost every I/O card and other device is plug-and-play....like since the Yggdrissl days.
system running, like a disk on usb.
I think you mean hot swapping? -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Fred A. Miller wrote:
Sorry I didn't "report" back on this subject sooner.....been VERY busy! Someone else here said that Linux doesn't handle plug and play ESATA nor SATA yet, and that is quite right. It's obviously an area that needs to
I'm running on a laptop that I purchased last summer. The original install was SuSE 10.0 (released early spring, 2006). This is an HP dv8000 system with two internal SATA drives. What do you mean Linux doesn't support SATA? I don't know about eSATA support, but claiming that Linux doesn't support SATA is just a flat out lie. It makes me think that you're not actually interested in help, but instead, are someone working for Microsoft or one of their partners and just posting propaganda FUD to scare people away from Linux. How much are you being paid to post that garbage?
be addressed SOON as more and more new boxen are being produced with ESATA connections.
You have to restart the system with power to the external hard drive and the ESATA port connected for it to work. Once you disconnect the drive, you'll get a "smart" error telling you that the drive is about to fail......annoying but that's about it.
Fred
-- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 The Friday 2007-12-07 at 05:00 -0500, Aaron Kulkis wrote:
Fred A. Miller wrote:
Sorry I didn't "report" back on this subject sooner.....been VERY busy! Someone else here said that Linux doesn't handle plug and play ESATA nor SATA yet, and that is quite right. It's obviously an area that needs to
I'm running on a laptop that I purchased last summer. The original install was SuSE 10.0 (released early spring, 2006). This is an HP dv8000 system with two internal SATA drives.
What do you mean Linux doesn't support SATA?
I don't know about eSATA support, but claiming that Linux doesn't support SATA is just a flat out lie.
Why do you say that? Why don't you read first the whole thread instead of jumping to rash conclusions about what the OP said? He asks about hot-swapping SATA or ESATA disks, that's very different from normal SATA support. He simply used the wrong name for it. - -- Cheers, Carlos E. R. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.4-svn0 (GNU/Linux) iD8DBQFHWYTStTMYHG2NR9URAsDHAKCWUNPvs4jM8xJML3S6s0QRasbfCgCeLWzy ZQ3ZbitHQzHI1WbOwnrwTBI= =z8Ea -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Carlos E. R. wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
The Friday 2007-12-07 at 05:00 -0500, Aaron Kulkis wrote:
Fred A. Miller wrote:
Sorry I didn't "report" back on this subject sooner.....been VERY busy! Someone else here said that Linux doesn't handle plug and play ESATA nor SATA yet, and that is quite right. It's obviously an area that needs to
I'm running on a laptop that I purchased last summer. The original install was SuSE 10.0 (released early spring, 2006). This is an HP dv8000 system with two internal SATA drives.
What do you mean Linux doesn't support SATA?
I don't know about eSATA support, but claiming that Linux doesn't support SATA is just a flat out lie.
Why do you say that? Why don't you read first the whole thread instead of jumping to rash conclusions about what the OP said?
That's the only part of the thread that had arrived when I wrote my reply.
He asks about hot-swapping SATA or ESATA disks, that's very different from normal SATA support. He simply used the wrong name for it.
I never claimed to be a mindreader (respond to what he means, not what he wrote) nor a psychic about what new messages would will say when they arrive in the future. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Dec 6, 2007 9:46 PM, Fred A. Miller
Sorry I didn't "report" back on this subject sooner.....been VERY busy! Someone else here said that Linux doesn't handle plug and play ESATA nor SATA yet, and that is quite right. It's obviously an area that needs to be addressed SOON as more and more new boxen are being produced with ESATA connections.
You have to restart the system with power to the external hard drive and the ESATA port connected for it to work. Once you disconnect the drive, you'll get a "smart" error telling you that the drive is about to fail......annoying but that's about it.
Fred
Fred, What exactly is your complaint? Because 10.3 should have pretty complete hot-swap capability. Or is a lack of automation your complant. ie. this works now: plug / manually mount / do work / manually unmount / unplug Repeat (possibly with a new /dev/sdX name even for the same drive). The issue I would like to see solved is hot swap of eSata drives in a Raid array with auto detection and rebuild etc. Currently the /dev/sdX name changes and the raid logic does not understand that a failed drive was replaced. Greg -- Greg Freemyer Litigation Triage Solutions Specialist http://www.linkedin.com/in/gregfreemyer First 99 Days Litigation White Paper - http://www.norcrossgroup.com/forms/whitepapers/99%20Days%20whitepaper.pdf The Norcross Group The Intersection of Evidence & Technology http://www.norcrossgroup.com -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Thu, 06 Dec 2007 21:46:30 -0500, Fred A. Miller wrote:
Someone else here said that Linux doesn't handle plug and play ESATA nor SATA yet, and that is quite right. It's obviously an area that needs to be addressed SOON as more and more new boxen are being produced with ESATA connections.
Oh come on Fred! It's far less serious then you make it to be.
You have to restart the system with power to the external hard drive and the ESATA port connected for it to work.
No! You just need to initiate a rescan which, AFAIK, you can do without a reboot.
Once you disconnect the drive, you'll get a "smart" error telling you that the drive is about to fail......annoying but that's about it.
Yes, because you also have to initiate a bus rescan for the system to notice that the drive has gone. Philipp -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Dec 9, 2007 10:57 AM, Philipp Thomas
On Thu, 06 Dec 2007 21:46:30 -0500, Fred A. Miller wrote:
Someone else here said that Linux doesn't handle plug and play ESATA nor SATA yet, and that is quite right. It's obviously an area that needs to be addressed SOON as more and more new boxen are being produced with ESATA connections.
Oh come on Fred! It's far less serious then you make it to be.
You have to restart the system with power to the external hard drive and the ESATA port connected for it to work.
No! You just need to initiate a rescan which, AFAIK, you can do without a reboot.
Once you disconnect the drive, you'll get a "smart" error telling you that the drive is about to fail......annoying but that's about it.
Yes, because you also have to initiate a bus rescan for the system to notice that the drive has gone.
Philipp
Auto rescan is supposed to happen now for many drivers. To check your hardware's support, see http://linux-ata.org/driver-status.html#matrix (this is actually 6-9 months old) And to manually rescan: echo - - - > /sys/class/scsi_host/hostX/scan Remember, sata drives are handled by the scsi subystem, so you're actually invoking a scsi rescan. Greg -- Greg Freemyer Litigation Triage Solutions Specialist http://www.linkedin.com/in/gregfreemyer First 99 Days Litigation White Paper - http://www.norcrossgroup.com/forms/whitepapers/99%20Days%20whitepaper.pdf The Norcross Group The Intersection of Evidence & Technology http://www.norcrossgroup.com -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
* Greg Freemyer (greg.freemyer@gmail.com) [20071209 17:15]:
Auto rescan is supposed to happen now for many drivers.
To check your hardware's support, see http://linux-ata.org/driver-status.html#matrix (this is actually 6-9 months old)
According to that matrix, auto rescan should work with the ICH7 on my mobo but fact is it doesn't with the eSATA drive I used for testing.
And to manually rescan:
echo - - - > /sys/class/scsi_host/hostX/scan
Remember, sata drives are handled by the scsi subystem, so you're actually invoking a scsi rescan.
I know that, I jsut forgot to mention the actual command. Thanks for providing that! Philipp -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
participants (12)
-
Aaron Kulkis -
Billie Walsh -
Carlos E. R. -
Felix Miata -
Fred A. Miller -
Greg Freemyer -
Greg KH -
Joe Morris (NTM) -
Morten Bjørnsvik -
Philipp Thomas -
Philipp Thomas -
Roger Oberholtzer