[opensuse] systemd and filesystems
I am running 13.2 (x86-64). I have several (removal) disks in the box. I manually unmount one and install another in its place to mount and use temporarily. When done I will put the original disk back in and remount it manually. I get this from systemd complaining about the old disk no longer being there. systemd[1]: Timed out waiting for device ST3160815AS. systemd[1]: Dependency failed for "blablabla". Over and over and over for every partition on it that is defined in the fstab file. What is this? My fstab has "ext4 defaults" for all the partitions on the disk. I think I read a thread here about a similar issue some time back but all I don't think I ever saw a real resolution. It had never affected me at the time. Not sure why, but it sure is now. What is the proper "systemd" way of doing this sort of thing now? Please don't tell me that this is no longer allowed. Thanks Mark -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On Fri, Jan 9, 2015 at 4:15 PM, Mark Hounschell <markh@compro.net> wrote:
I am running 13.2 (x86-64). I have several (removal) disks in the box. I manually unmount one and install another in its place to mount and use temporarily. When done I will put the original disk back in and remount it manually. I get this from systemd complaining about the old disk no longer being there.
What is "old" and what is "new" here? But it sounds like you hit https://bugzilla.opensuse.org/show_bug.cgi?id=902612.
systemd[1]: Timed out waiting for device ST3160815AS. systemd[1]: Dependency failed for "blablabla".
Over and over and over for every partition on it that is defined in the fstab file.
What is this? My fstab has "ext4 defaults" for all the partitions on the disk. I think I read a thread here about a similar issue some time back but all I don't think I ever saw a real resolution. It had never affected me at the time. Not sure why, but it sure is now. What is the proper "systemd" way of doing this sort of thing now? Please don't tell me that this is no longer allowed.
Thanks Mark
-- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
-- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On Fri, 09 Jan 2015 08:15:09 -0500 Mark Hounschell wrote:
I am running 13.2 (x86-64). I have several (removal) disks in the box. I manually unmount one and install another in its place to mount and use temporarily. When done I will put the original disk back in and remount it manually.
Hi Mark, It would help if you could provide more details about this arrangement and how you're taking advantage of it at the command line. I had some difficulty interpreting this sentence: "I have several (removal) disks in the box." My first guess says you meant to type "removable." I then looked up the specs for the hard drive model number you supplied, so at least I know it's SATA (and 'enterprise class,' if it's hot swappable.) You seem to be saying that you have a 'hot swap' SATA drive bay installed in the computer and that you're using it with a number of these drives. Is this correct? If so, please post the related entries you've created in /etc/fstab Another set of clues that would be extremely helpful would be actual examples of the commands you're using to mount and dismount the partitions. Your examples should cover the entire sequence of events, from what you do immediately after you've plugged in the first drive, all the way through to the point where you've swapped them out and you're seeing the errors. regards, Carl -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On 01/12/2015 07:56 AM, Carl Hartung wrote:
On Fri, 09 Jan 2015 08:15:09 -0500 Mark Hounschell wrote:
I am running 13.2 (x86-64). I have several (removal) disks in the box. I manually unmount one and install another in its place to mount and use temporarily. When done I will put the original disk back in and remount it manually.
Hi Mark,
It would help if you could provide more details about this arrangement and how you're taking advantage of it at the command line. I had some difficulty interpreting this sentence:
"I have several (removal) disks in the box."
My first guess says you meant to type "removable." I then looked up the specs for the hard drive model number you supplied, so at least I know it's SATA (and 'enterprise class,' if it's hot swappable.)
Yes, "removable".
You seem to be saying that you have a 'hot swap' SATA drive bay installed in the computer and that you're using it with a number of these drives. Is this correct? If so, please post the related entries you've created in /etc/fstab
Another set of clues that would be extremely helpful would be actual examples of the commands you're using to mount and dismount the partitions. Your examples should cover the entire sequence of events, from what you do immediately after you've plugged in the first drive, all the way through to the point where you've swapped them out and you're seeing the errors.
Attempting to re-create the problem so that I can give more info resulted in NOT being able to reproduce the issue originally reported. Anyway this is what I did to try to re-create it. First, I boot the machine with all disks defined in fstab installed. The particular disk I that I will unmount and replace today has this fstab entry: /dev/disk/by-id/ata-ST3500320NS_9QM1N3A5-part1 /lcrs ext3 defaults 1 2 After it's boot is complete, I login and umount /lcrs as root via #umount /lcrs After insuring /lcrs is in fact not mounted any longer, I then turn off the power to that drive bay and remove it. I wait several minutes for the messages but today, I get none. I then install into the bay a different drive and power it up. At this point "dmesg" shows that the new disk was detected and it has the same /dev/sd# number as the original disk. So I mount it via #mount /dev/sdf1 /mnt Then I wait again for the messages I saw on Friday..... Today, none appear. All seems OK. I wait a several more minutes still all is OK. So I then unmount the disk via #umount /mnt Then turn the power to the bay off and remove my disk. I then install and power up the original disk and after the system detects this it is automatically remounted as /lcrs. All is good today. I am unable to duplicate what was happening on Friday. Regards Mark -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On Fri, 9 Jan 2015 14:15, Mark Hounschell <markh@...> wrote:
I am running 13.2 (x86-64). I have several (removal) disks in the box. I manually unmount one and install another in its place to mount and use temporarily. When done I will put the original disk back in and remount it manually. I get this from systemd complaining about the old disk no longer being there.
systemd[1]: Timed out waiting for device ST3160815AS. systemd[1]: Dependency failed for "blablabla".
Over and over and over for every partition on it that is defined in the fstab file.
What is this? My fstab has "ext4 defaults" for all the partitions on the disk. I think I read a thread here about a similar issue some time back but all I don't think I ever saw a real resolution. It had never affected me at the time. Not sure why, but it sure is now. What is the proper "systemd" way of doing this sort of thing now? Please don't tell me that this is no longer allowed.
Have a look at the 'nofail' mount / fstab option. "man 8 mount" Here a 'defauls,nofail' would be helpfull. -Yamaban. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On 01/09/2015 02:15 PM, Mark Hounschell wrote:
I am running 13.2 (x86-64). I have several (removal) disks in the box. I manually unmount one and install another in its place to mount and use temporarily. When done I will put the original disk back in and remount it manually. I get this from systemd complaining about the old disk no longer being there.
systemd[1]: Timed out waiting for device ST3160815AS. systemd[1]: Dependency failed for "blablabla".
Over and over and over for every partition on it that is defined in the fstab file.
What is this? My fstab has "ext4 defaults" for all the partitions on the disk. I think I read a thread here about a similar issue some time back but all I don't think I ever saw a real resolution. It had never affected me at the time. Not sure why, but it sure is now. What is the proper "systemd" way of doing this sort of thing now? Please don't tell me that this is no longer allowed.
Thanks Mark
Systenmd also is able to replace fstab - or parts of it. On systemd linux systems nowadays you sometimes only see real devices in /etc/fstab and a unit generator "systemd-fstab-generator" generates systemd units from the fstab. To see the filesystems systemd controls, see systemctl -a -t mount Another feature of systemd are dependencies. So if systemd wants to mount a filesystem, it can wait for a device to appear before. Use systemd list-dependencies to inspect this. I think you have to cleanup something, that came from installation time or use systemctl disable blabla.mount instead of simply umount. Just a guess, i did not try ...
On Tue, 13 Jan 2015, Florian Gleixner wrote:
On 01/09/2015 02:15 PM, Mark Hounschell wrote:
I am running 13.2 (x86-64). I have several (removal) disks in the box. I manually unmount one and install another in its place to mount and use temporarily. When done I will put the original disk back in and remount it manually. I get this from systemd complaining about the old disk no longer being there.
systemd[1]: Timed out waiting for device ST3160815AS. systemd[1]: Dependency failed for "blablabla".
Over and over and over for every partition on it that is defined in the fstab file.
What is this? My fstab has "ext4 defaults" for all the partitions on the disk. I think I read a thread here about a similar issue some time back but all I don't think I ever saw a real resolution. It had never affected me at the time. Not sure why, but it sure is now. What is the proper "systemd" way of doing this sort of thing now? Please don't tell me that this is no longer allowed.
Thanks Mark
Systenmd also is able to replace fstab - or parts of it. On systemd linux systems nowadays you sometimes only see real devices in /etc/fstab and a unit generator "systemd-fstab-generator" generates systemd units from the fstab. To see the filesystems systemd controls, see systemctl -a -t mount
Another feature of systemd are dependencies. So if systemd wants to mount a filesystem, it can wait for a device to appear before. Use
systemd list-dependencies
to inspect this.
I think you have to cleanup something, that came from installation time or use
systemctl disable blabla.mount
instead of simply umount. Just a guess, i did not try ...
Based on my own experiences, this might be the issue. I did a umount on a disk listed in /etc/fstab do do some maintenance on it, while I was working on it systemMd noted it was not longer mounted and remounted it for me. So trying to mount/unmount disks that are listed in /etc/fstab is dangerous without first informing systemMd either by the above mentioned disabled or perhaps a noauto in /etc/fstab. Very unexpected behaviour for old timers. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On 01/14/2015 04:47 AM, Michael Hamilton wrote:
On Tue, 13 Jan 2015, Florian Gleixner wrote:
On 01/09/2015 02:15 PM, Mark Hounschell wrote:
I am running 13.2 (x86-64). I have several (removal) disks in the box. I manually unmount one and install another in its place to mount and use temporarily. When done I will put the original disk back in and remount it manually. I get this from systemd complaining about the old disk no longer being there.
systemd[1]: Timed out waiting for device ST3160815AS. systemd[1]: Dependency failed for "blablabla".
Over and over and over for every partition on it that is defined in the fstab file.
What is this? My fstab has "ext4 defaults" for all the partitions on the disk. I think I read a thread here about a similar issue some time back but all I don't think I ever saw a real resolution. It had never affected me at the time. Not sure why, but it sure is now. What is the proper "systemd" way of doing this sort of thing now? Please don't tell me that this is no longer allowed.
Thanks Mark
Systenmd also is able to replace fstab - or parts of it. On systemd linux systems nowadays you sometimes only see real devices in /etc/fstab and a unit generator "systemd-fstab-generator" generates systemd units from the fstab. To see the filesystems systemd controls, see systemctl -a -t mount
Another feature of systemd are dependencies. So if systemd wants to mount a filesystem, it can wait for a device to appear before. Use
systemd list-dependencies
to inspect this.
I think you have to cleanup something, that came from installation time or use
systemctl disable blabla.mount
instead of simply umount. Just a guess, i did not try ...
Based on my own experiences, this might be the issue. I did a umount on a disk listed in /etc/fstab do do some maintenance on it, while I was working on it systemMd noted it was not longer mounted and remounted it for me. So trying to mount/unmount disks that are listed in /etc/fstab is dangerous without first informing systemMd either by the above mentioned disabled or perhaps a noauto in /etc/fstab.
Very unexpected behaviour for old timers.
What I found was that systemd would not attempt to remount it unless it was actually removed and then reinstalled. Or powered off then on. I'm thinking that my original problem might have been that I may have inadvertently powered down then up then down again. Systemd seems only to attempt a remount if it is re-attached or powered up. Simply unmounting then powering down of the disk causes systemd to do nothing. Which is good. I can live with the remount after powering it back up. Unexpected yes, but OK with me. Regards Mark -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On 2015-01-14 14:01, Mark Hounschell wrote:
On 01/14/2015 04:47 AM, Michael Hamilton wrote:
On Tue, 13 Jan 2015, Florian Gleixner wrote:
On 01/09/2015 02:15 PM, Mark Hounschell wrote:
What I found was that systemd would not attempt to remount it unless it was actually removed and then reinstalled. Or powered off then on. I'm thinking that my original problem might have been that I may have inadvertently powered down then up then down again. Systemd seems only to attempt a remount if it is re-attached or powered up.
Simply unmounting then powering down of the disk causes systemd to do nothing. Which is good. I can live with the remount after powering it back up. Unexpected yes, but OK with me.
I had problems with systemd remounting while I was reformatting a partition. Or worse, creating/removing partitions. -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" at Telcontar)
On Thu, 15 Jan 2015, Carlos E. R. wrote:
On 2015-01-14 14:01, Mark Hounschell wrote:
On 01/14/2015 04:47 AM, Michael Hamilton wrote:
On Tue, 13 Jan 2015, Florian Gleixner wrote:
On 01/09/2015 02:15 PM, Mark Hounschell wrote:
What I found was that systemd would not attempt to remount it unless it was actually removed and then reinstalled. Or powered off then on. I'm thinking that my original problem might have been that I may have inadvertently powered down then up then down again. Systemd seems only to attempt a remount if it is re-attached or powered up.
Simply unmounting then powering down of the disk causes systemd to do nothing. Which is good. I can live with the remount after powering it back up. Unexpected yes, but OK with me.
I had problems with systemd remounting while I was reformatting a partition. Or worse, creating/removing partitions.
I was altering the partitioning and re-sizing a filesystem. Perhaps such alterations prompt systemd to reinspect the state of play and take action, -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On 2015-01-14 19:51, Michael Hamilton wrote:
On Thu, 15 Jan 2015, Carlos E. R. wrote:
> I was altering the partitioning and re-sizing a filesystem. Perhaps such
alterations prompt systemd to reinspect the state of play and take action,
Yes, I think so, because new filesystems appear. And of course, during this type of maintenance, you may not have yet adjusted fstab for the new situation. First format, then edit... -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" at Telcontar)
participants (7)
-
Andrei Borzenkov -
Carl Hartung -
Carlos E. R. -
Florian Gleixner -
Mark Hounschell -
Michael Hamilton -
Yamaban