On 11/24/2017 9:41 AM, Christian Betz wrote:
> First, a potential non-answer: Use containers. Or Flatpak. CL supports both
this is in part a real answer. some things are really done better as a container
or a flatpak.... not everything of course.
>
> (a) you have some device that can run clear linux and you want to package
> your own stuff on it to make an "appliance" which can be remotely updated
> using swupd. for something concrete: imagine i wanted to make and brand a
> digital signage controller based on intel NUCs. it should magically update
> itself with my branded software by contacting a swupd server that I host.
as we're expanding our focus a bit in the IOT kind of space (big IOT, not lightbulbs),
this indeed one of our models
>
> for something like this, i think i'm interested in what CL calls "Mixer":
>
> https://clearlinux.org/features/mixer
>
> https://clearlinux.org/documentation/clear-linux/guides/maintenance/mixer.h…
and mixer is exactly for these kind of scenarios; be it a device maker for their own
devices, or a SRE team for a datacenter managing a set of systems, mixer is how
you get full control, not just of what you add, but also when and how and what you update
> BTW: CL specifically uses something called 'autospec' to generate spec
> files for RPM builds more easily (see:
> https://github.com/clearlinux/autospec, https://github.com/clearlinux-pkgs).
> AFAICT not all CL packages are generated using autospec; in other words I
> don't think you *have* to use autospec but you are encouraged to.
autospec for us is targeted for "95%", the packages that have sane (well, "standardized",
lets not debate if autoconf or cmake or .. are sane or insane) build systems.
In addition, autospec is basically our packaging standard; it's code not english, but
the way autospec puts down packages is by construction how we want to package.
>
> (c) you are a developer of some product or open source project and you want
> to make available for purchase/download your application for clear linux
> (i.e. for other users of clear linux). In a "classic distro" model you
> might make available deb or RPM packages for popular distros. Or better:
> you have repos for yum/app. But what should you do for users of Clear
> Linux? If you cannot or don't want the CL team to officially adopt your
> package? I cannot answer this one :(
>
if we miss open source stuff, just ask for it to be included, by means
of an URL to the release tar file....
if it just autospecs and it's not insane in terms of the legal/security side, we'll
just add it.
(rule of thumb: if it's good enough for fedora and debian it's likely not insane)
Hi,
Disclaimer: I am just a new user, only three installations deep. But I'll
try to answer based on what I know from the docs and my experience.
Corrections/clarifications are appreciated!
First, a potential non-answer: Use containers. Or Flatpak. CL supports both
very well. Docker works and I have been able to run Firefox+GNOME/Wayland
by using a nightly FF wayland-supporting flatpak build. The Clear
Containers runtime (cc-runtime) and qemu-lite used for running containers
in VMs is the reason many of are trying out CL. So we shouldn't feel bad
about using containers :)
Beyond that I have a few of my own scenarios in mind that could be similar
to Alessio's:
(a) you have some device that can run clear linux and you want to package
your own stuff on it to make an "appliance" which can be remotely updated
using swupd. for something concrete: imagine i wanted to make and brand a
digital signage controller based on intel NUCs. it should magically update
itself with my branded software by contacting a swupd server that I host.
for something like this, i think i'm interested in what CL calls "Mixer":
https://clearlinux.org/features/mixerhttps://clearlinux.org/documentation/clear-linux/guides/maintenance/mixer.h…
(b) in our datacenter at $DAYJOB we we build dozens of custom RPM packages
for CentOS 7 for various services and configurations, mostly based on
$DAYJOB_PROGRAMMING_LANGUAGE. switching to CL+swupdate but continuing
writing custom RPM specs and linking them into bundles based on the mixer
procedures linked above is a good fit for us.
BTW: CL specifically uses something called 'autospec' to generate spec
files for RPM builds more easily (see:
https://github.com/clearlinux/autospec, https://github.com/clearlinux-pkgs).
AFAICT not all CL packages are generated using autospec; in other words I
don't think you *have* to use autospec but you are encouraged to.
(c) you are a developer of some product or open source project and you want
to make available for purchase/download your application for clear linux
(i.e. for other users of clear linux). In a "classic distro" model you
might make available deb or RPM packages for popular distros. Or better:
you have repos for yum/app. But what should you do for users of Clear
Linux? If you cannot or don't want the CL team to officially adopt your
package? I cannot answer this one :(
On Fri, Nov 24, 2017 at 5:07 AM, Victor Rodriguez <vm.rod25(a)gmail.com>
wrote:
> On Nov 24, 2017 2:17 AM, "Alessio Igor Bogani" <
> alessioigorbogani(a)gmail.com>
> wrote:
>
> Hi all,
>
> Can someone explain me how is it supposed that an developer ships his
> own software and handle related build and runtime dependencies in
> ClearLinux?
>
>
>
> Hi
>
> What do you mean by related runtime dependencies? Can you provide a
> detailed example of what software you want to ship?
>
> Regards
>
> Victor
>
> Thanks a lot!
>
> Ciao,
> Alessio
> _______________________________________________
> Dev mailing list
> Dev(a)lists.clearlinux.org
> https://lists.clearlinux.org/mailman/listinfo/dev
> _______________________________________________
> Dev mailing list
> Dev(a)lists.clearlinux.org
> https://lists.clearlinux.org/mailman/listinfo/dev
>
--
"the new garbage collector will be an arena-based, quad-color incremental,
generational, non-copying, high-speed, cache-optimized garbage collector"
-- LuaJIT Roadmap
On Nov 24, 2017 2:17 AM, "Alessio Igor Bogani" <alessioigorbogani(a)gmail.com>
wrote:
Hi all,
Can someone explain me how is it supposed that an developer ships his
own software and handle related build and runtime dependencies in
ClearLinux?
Hi
What do you mean by related runtime dependencies? Can you provide a
detailed example of what software you want to ship?
Regards
Victor
Thanks a lot!
Ciao,
Alessio
_______________________________________________
Dev mailing list
Dev(a)lists.clearlinux.orghttps://lists.clearlinux.org/mailman/listinfo/dev
Hi all,
Can someone explain me how is it supposed that an developer ships his
own software and handle related build and runtime dependencies in
ClearLinux?
Thanks a lot!
Ciao,
Alessio
Clear doesn't have this run by default. You could try making a service
(Type=oneshot) to run the device scan and make the mount unit require
that service and see if that gets the uuid detection.
On Mon, Nov 20, 2017 at 3:16 PM, Richard Thornton
<richie.thornton(a)gmail.com> wrote:
> How does CL implement the btrfs device scan?
>
> https://btrfs.wiki.kernel.org/index.php/Problem_FAQ
>
> Then you need to ensure that you run a btrfs device scan first:
>
> # btrfs device scan
>
> This should be in many distributions' startup scripts (and initrd
> images, if your root filesystem is btrfs), but you may have to add it
> yourself.
>
> On 21 November 2017 at 10:09, Richard Thornton
> <richie.thornton(a)gmail.com> wrote:
>> I can see:
>>
>> [ 20.062217] mpt2sas_cm0: LSISAS2008: FWVersion(07.15.04.00),
>> ChipRevision(0x03), BiosVersion(00.00.00.00)
>> ...
>> [ 20.198645] Btrfs loaded, crc32c=crc32c-intel
>> ...
>> [ 30.073488] BTRFS: device fsid a4d1ef9d-e75f-41dd-acac-9aaa79428965
>> devid 2 transid 16 /dev/sdb
>> [ 30.073911] BTRFS info (device sdb): disk space caching is enabled
>> [ 30.073912] BTRFS info (device sdb): has skinny extents
>> [ 30.074170] BTRFS warning (device sdb): devid 1 uuid
>> fc680ff9-c623-4baf-85d4-c634d2dbc6ef is missing
>> [ 30.074171] BTRFS error (device sdb): failed to read the system array: -5
>> [ 30.101015] BTRFS error (device sdb): open_ctree failed
>>
>> It's interesting that it finds "devid 2" but errors on "devid 1" missing?
>>
>> On 21 November 2017 at 08:54, Richard Thornton
>> <richie.thornton(a)gmail.com> wrote:
>>> Thanks, what would be the preferred way to do this after the CL system had
>>> booted:
>>>
>>> Start mount
>>> Start nmbd
>>> Start Smdb
>>> Start docker (plex)
>>>
>>>
>>> On 21 Nov. 2017 08:25, "Douglas, William" <william.douglas(a)intel.com> wrote:
>>>
>>> You might be seeing an instance where /dev/disk/by-uuid is not fully
>>> populated by the time your service runs. Having your disk attached to
>>> an adapter makes this more likely to me.
>>>
>>> Maybe try
>>> https://www.freedesktop.org/software/systemd/man/systemd.mount.html#Timeout…
>>> but that's just for the mount command so I'm not sure if it will even
>>> attempt to run the mount or not based on the uuid missing log you had
>>> but that's the only thing I could find that might work.
>>>
>>> On Mon, Nov 20, 2017 at 1:08 PM, Richard Thornton
>>> <richie.thornton(a)gmail.com> wrote:
>>>> I should add that the btrfs disks are attached to an LSI pcie adapter.
>>>>
>>>> On 21 Nov. 2017 07:56, "Richard Thornton" <richie.thornton(a)gmail.com>
>>>> wrote:
>>>>>
>>>>> My dir:
>>>>>
>>>>> root(a)thor ~ # ls /etc/systemd/system/
>>>>>
>>>>> multi-user.target.wants nmbd.service smbd.service storage.mount
>>>>>
>>>>> On 21 November 2017 at 07:42, Richard Thornton
>>>>> <richie.thornton(a)gmail.com> wrote:
>>>>> > dmesg snippet:
>>>>> >
>>>>> > [ 30.073488] BTRFS: device fsid a4d1ef9d-e75f-41dd-acac-9aaa79428965
>>>>> > devid 2 transid 16 /dev/sdb
>>>>> > [ 30.073911] BTRFS info (device sdb): disk space caching is enabled
>>>>> > [ 30.073912] BTRFS info (device sdb): has skinny extents
>>>>> > [ 30.074170] BTRFS warning (device sdb): devid 1 uuid
>>>>> > fc680ff9-c623-4baf-85d4-c634d2dbc6ef is missing
>>>>> > [ 30.074171] BTRFS error (device sdb): failed to read the system
>>>>> > array: -5
>>>>> > [ 30.101015] BTRFS error (device sdb): open_ctree failed
>>>>> >
>>>>> > On 21 November 2017 at 07:21, Richard Thornton
>>>>> > <richie.thornton(a)gmail.com> wrote:
>>>>> >> Thanks, the below works with "systemctl start storage.mount" but not
>>>>> >> when enabled for boot:
>>>>> >>
>>>>> >> [Unit]
>>>>> >> Description=Mount
>>>>> >> Before=nmbd.service
>>>>> >>
>>>>> >> [Mount]
>>>>> >> What=/dev/disk/by-uuid/a4d1ef9d-e75f-41dd-acac-9aaa79428965
>>>>> >> Where=/storage
>>>>> >>
>>>>> >> [Install]
>>>>> >> WantedBy=multi-user.target
>>>>> >>
>>>>> >> On 21 November 2017 at 06:04, Douglas, William
>>>>> >> <william.douglas(a)intel.com> wrote:
>>>>> >>> So I didn't see any issues when I did this. Can you show the
>>>>> >>> storage.mount file and tell me what is the target that requires
>>>>> >>> storage.mount?
>>>>> >>>
>>>>> >>> On Mon, Nov 20, 2017 at 8:49 AM, Douglas, William
>>>>> >>> <william.douglas(a)intel.com> wrote:
>>>>> >>>> I haven't tried this recently so I'll give it a go today and get
>>>>> >>>> back
>>>>> >>>> to
>>>>> >>>> you.
>>>>> >>>>
>>>>> >>>>
>>>>> >>>> On Nov 20, 2017 02:45, "Richard Thornton"
>>>>> >>>> <richie.thornton(a)gmail.com>
>>>>> >>>> wrote:
>>>>> >>>>
>>>>> >>>> Thanks, I finally got around to setting this up, systemd is loading
>>>>> >>>> samba on boot and I could mount btrfs with
>>>>> >>>> "systemctl start storage.mount" after boot.
>>>>> >>>>
>>>>> >>>> The issue I'm having is I can't get a btrfs to mount on boot, I
>>>>> >>>> tried
>>>>> >>>> systemd and fstab. I think it could be something to do with this:
>>>>> >>>>
>>>>> >>>>
>>>>> >>>>
>>>>> >>>> https://blogs.gentoo.org/remi/2014/12/05/gentoo-btrfs-arrays-and-systemd-a-…
>>>>> >>>>
>>>>> >>>> I spent hours on trying to get this to persistent mount, including
>>>>> >>>> doing what the above link stated, can someone please help me out?
>>>>> >>>>
>>>>> >>>> Thanks.
>>>>> >>>>
>>>>> >>>> Richard
>>>>> >>>>
>>>>> >>>> On 20 October 2017 at 02:19, Nesius, Robert A
>>>>> >>>> <robert.a.nesius(a)intel.com>
>>>>> >>>> wrote:
>>>>> >>>>> Yep - you got it.
>>>>> >>>>>
>>>>> >>>>> Technically systemd supports /etc/fstab for backwards compatibility
>>>>> >>>>> - it
>>>>> >>>>> generates a mount unit for each line in the file.
>>>>> >>>>>
>>>>> >>>>> -Rob
>>>>> >>>>>
>>>>> >>>>> Sent from mobile device.
>>>>> >>>>>
>>>>> >>>>> On Oct 18, 2017, at 9:03 PM, Richard Thornton
>>>>> >>>>> <richie.thornton(a)gmail.com>
>>>>> >>>>> wrote:
>>>>> >>>>>
>>>>> >>>>> Hi,
>>>>> >>>>>
>>>>> >>>>> Sorry to bother you all.
>>>>> >>>>>
>>>>> >>>>> I like Clear Linux a lot and Arjan has been a great help.
>>>>> >>>>>
>>>>> >>>>> Posted to dev@ because I couldn't see a user@
>>>>> >>>>>
>>>>> >>>>> I'm going to use Clear Linux for a 24/7 home server, I like the
>>>>> >>>>> upgrade system (used to use Ubuntu).
>>>>> >>>>>
>>>>> >>>>> I'm going to use Docker containers for my applications.
>>>>> >>>>>
>>>>> >>>>> I want to setup some persistence, if there is a reboot I would like
>>>>> >>>>> my
>>>>> >>>>> storage disks to mount, Samba to start and my Docker containers to
>>>>> >>>>> run.
>>>>> >>>>>
>>>>> >>>>> I think the way to do this is through systemd on CL, if so, I need
>>>>> >>>>> to
>>>>> >>>>> create the /etc/systemd structure?
>>>>> >>>>>
>>>>> >>>>> Just looking for best practices, advice and any documentation that
>>>>> >>>>> may
>>>>> >>>>> be out there.
>>>>> >>>>>
>>>>> >>>>> Thanks.
>>>>> >>>>>
>>>>> >>>>> Richard
>>>>> >>>>> _______________________________________________
>>>>> >>>>> Dev mailing list
>>>>> >>>>> Dev(a)lists.clearlinux.org
>>>>> >>>>> https://lists.clearlinux.org/mailman/listinfo/dev
>>>>> >>>> _______________________________________________
>>>>> >>>> Dev mailing list
>>>>> >>>> Dev(a)lists.clearlinux.org
>>>>> >>>> https://lists.clearlinux.org/mailman/listinfo/dev
>>>>> >>>>
>>>>> >>>>
>>>
>>>
How does CL implement the btrfs device scan?
https://btrfs.wiki.kernel.org/index.php/Problem_FAQ
Then you need to ensure that you run a btrfs device scan first:
# btrfs device scan
This should be in many distributions' startup scripts (and initrd
images, if your root filesystem is btrfs), but you may have to add it
yourself.
On 21 November 2017 at 10:09, Richard Thornton
<richie.thornton(a)gmail.com> wrote:
> I can see:
>
> [ 20.062217] mpt2sas_cm0: LSISAS2008: FWVersion(07.15.04.00),
> ChipRevision(0x03), BiosVersion(00.00.00.00)
> ...
> [ 20.198645] Btrfs loaded, crc32c=crc32c-intel
> ...
> [ 30.073488] BTRFS: device fsid a4d1ef9d-e75f-41dd-acac-9aaa79428965
> devid 2 transid 16 /dev/sdb
> [ 30.073911] BTRFS info (device sdb): disk space caching is enabled
> [ 30.073912] BTRFS info (device sdb): has skinny extents
> [ 30.074170] BTRFS warning (device sdb): devid 1 uuid
> fc680ff9-c623-4baf-85d4-c634d2dbc6ef is missing
> [ 30.074171] BTRFS error (device sdb): failed to read the system array: -5
> [ 30.101015] BTRFS error (device sdb): open_ctree failed
>
> It's interesting that it finds "devid 2" but errors on "devid 1" missing?
>
> On 21 November 2017 at 08:54, Richard Thornton
> <richie.thornton(a)gmail.com> wrote:
>> Thanks, what would be the preferred way to do this after the CL system had
>> booted:
>>
>> Start mount
>> Start nmbd
>> Start Smdb
>> Start docker (plex)
>>
>>
>> On 21 Nov. 2017 08:25, "Douglas, William" <william.douglas(a)intel.com> wrote:
>>
>> You might be seeing an instance where /dev/disk/by-uuid is not fully
>> populated by the time your service runs. Having your disk attached to
>> an adapter makes this more likely to me.
>>
>> Maybe try
>> https://www.freedesktop.org/software/systemd/man/systemd.mount.html#Timeout…
>> but that's just for the mount command so I'm not sure if it will even
>> attempt to run the mount or not based on the uuid missing log you had
>> but that's the only thing I could find that might work.
>>
>> On Mon, Nov 20, 2017 at 1:08 PM, Richard Thornton
>> <richie.thornton(a)gmail.com> wrote:
>>> I should add that the btrfs disks are attached to an LSI pcie adapter.
>>>
>>> On 21 Nov. 2017 07:56, "Richard Thornton" <richie.thornton(a)gmail.com>
>>> wrote:
>>>>
>>>> My dir:
>>>>
>>>> root(a)thor ~ # ls /etc/systemd/system/
>>>>
>>>> multi-user.target.wants nmbd.service smbd.service storage.mount
>>>>
>>>> On 21 November 2017 at 07:42, Richard Thornton
>>>> <richie.thornton(a)gmail.com> wrote:
>>>> > dmesg snippet:
>>>> >
>>>> > [ 30.073488] BTRFS: device fsid a4d1ef9d-e75f-41dd-acac-9aaa79428965
>>>> > devid 2 transid 16 /dev/sdb
>>>> > [ 30.073911] BTRFS info (device sdb): disk space caching is enabled
>>>> > [ 30.073912] BTRFS info (device sdb): has skinny extents
>>>> > [ 30.074170] BTRFS warning (device sdb): devid 1 uuid
>>>> > fc680ff9-c623-4baf-85d4-c634d2dbc6ef is missing
>>>> > [ 30.074171] BTRFS error (device sdb): failed to read the system
>>>> > array: -5
>>>> > [ 30.101015] BTRFS error (device sdb): open_ctree failed
>>>> >
>>>> > On 21 November 2017 at 07:21, Richard Thornton
>>>> > <richie.thornton(a)gmail.com> wrote:
>>>> >> Thanks, the below works with "systemctl start storage.mount" but not
>>>> >> when enabled for boot:
>>>> >>
>>>> >> [Unit]
>>>> >> Description=Mount
>>>> >> Before=nmbd.service
>>>> >>
>>>> >> [Mount]
>>>> >> What=/dev/disk/by-uuid/a4d1ef9d-e75f-41dd-acac-9aaa79428965
>>>> >> Where=/storage
>>>> >>
>>>> >> [Install]
>>>> >> WantedBy=multi-user.target
>>>> >>
>>>> >> On 21 November 2017 at 06:04, Douglas, William
>>>> >> <william.douglas(a)intel.com> wrote:
>>>> >>> So I didn't see any issues when I did this. Can you show the
>>>> >>> storage.mount file and tell me what is the target that requires
>>>> >>> storage.mount?
>>>> >>>
>>>> >>> On Mon, Nov 20, 2017 at 8:49 AM, Douglas, William
>>>> >>> <william.douglas(a)intel.com> wrote:
>>>> >>>> I haven't tried this recently so I'll give it a go today and get
>>>> >>>> back
>>>> >>>> to
>>>> >>>> you.
>>>> >>>>
>>>> >>>>
>>>> >>>> On Nov 20, 2017 02:45, "Richard Thornton"
>>>> >>>> <richie.thornton(a)gmail.com>
>>>> >>>> wrote:
>>>> >>>>
>>>> >>>> Thanks, I finally got around to setting this up, systemd is loading
>>>> >>>> samba on boot and I could mount btrfs with
>>>> >>>> "systemctl start storage.mount" after boot.
>>>> >>>>
>>>> >>>> The issue I'm having is I can't get a btrfs to mount on boot, I
>>>> >>>> tried
>>>> >>>> systemd and fstab. I think it could be something to do with this:
>>>> >>>>
>>>> >>>>
>>>> >>>>
>>>> >>>> https://blogs.gentoo.org/remi/2014/12/05/gentoo-btrfs-arrays-and-systemd-a-…
>>>> >>>>
>>>> >>>> I spent hours on trying to get this to persistent mount, including
>>>> >>>> doing what the above link stated, can someone please help me out?
>>>> >>>>
>>>> >>>> Thanks.
>>>> >>>>
>>>> >>>> Richard
>>>> >>>>
>>>> >>>> On 20 October 2017 at 02:19, Nesius, Robert A
>>>> >>>> <robert.a.nesius(a)intel.com>
>>>> >>>> wrote:
>>>> >>>>> Yep - you got it.
>>>> >>>>>
>>>> >>>>> Technically systemd supports /etc/fstab for backwards compatibility
>>>> >>>>> - it
>>>> >>>>> generates a mount unit for each line in the file.
>>>> >>>>>
>>>> >>>>> -Rob
>>>> >>>>>
>>>> >>>>> Sent from mobile device.
>>>> >>>>>
>>>> >>>>> On Oct 18, 2017, at 9:03 PM, Richard Thornton
>>>> >>>>> <richie.thornton(a)gmail.com>
>>>> >>>>> wrote:
>>>> >>>>>
>>>> >>>>> Hi,
>>>> >>>>>
>>>> >>>>> Sorry to bother you all.
>>>> >>>>>
>>>> >>>>> I like Clear Linux a lot and Arjan has been a great help.
>>>> >>>>>
>>>> >>>>> Posted to dev@ because I couldn't see a user@
>>>> >>>>>
>>>> >>>>> I'm going to use Clear Linux for a 24/7 home server, I like the
>>>> >>>>> upgrade system (used to use Ubuntu).
>>>> >>>>>
>>>> >>>>> I'm going to use Docker containers for my applications.
>>>> >>>>>
>>>> >>>>> I want to setup some persistence, if there is a reboot I would like
>>>> >>>>> my
>>>> >>>>> storage disks to mount, Samba to start and my Docker containers to
>>>> >>>>> run.
>>>> >>>>>
>>>> >>>>> I think the way to do this is through systemd on CL, if so, I need
>>>> >>>>> to
>>>> >>>>> create the /etc/systemd structure?
>>>> >>>>>
>>>> >>>>> Just looking for best practices, advice and any documentation that
>>>> >>>>> may
>>>> >>>>> be out there.
>>>> >>>>>
>>>> >>>>> Thanks.
>>>> >>>>>
>>>> >>>>> Richard
>>>> >>>>> _______________________________________________
>>>> >>>>> Dev mailing list
>>>> >>>>> Dev(a)lists.clearlinux.org
>>>> >>>>> https://lists.clearlinux.org/mailman/listinfo/dev
>>>> >>>> _______________________________________________
>>>> >>>> Dev mailing list
>>>> >>>> Dev(a)lists.clearlinux.org
>>>> >>>> https://lists.clearlinux.org/mailman/listinfo/dev
>>>> >>>>
>>>> >>>>
>>
>>
I can see:
[ 20.062217] mpt2sas_cm0: LSISAS2008: FWVersion(07.15.04.00),
ChipRevision(0x03), BiosVersion(00.00.00.00)
...
[ 20.198645] Btrfs loaded, crc32c=crc32c-intel
...
[ 30.073488] BTRFS: device fsid a4d1ef9d-e75f-41dd-acac-9aaa79428965
devid 2 transid 16 /dev/sdb
[ 30.073911] BTRFS info (device sdb): disk space caching is enabled
[ 30.073912] BTRFS info (device sdb): has skinny extents
[ 30.074170] BTRFS warning (device sdb): devid 1 uuid
fc680ff9-c623-4baf-85d4-c634d2dbc6ef is missing
[ 30.074171] BTRFS error (device sdb): failed to read the system array: -5
[ 30.101015] BTRFS error (device sdb): open_ctree failed
It's interesting that it finds "devid 2" but errors on "devid 1" missing?
On 21 November 2017 at 08:54, Richard Thornton
<richie.thornton(a)gmail.com> wrote:
> Thanks, what would be the preferred way to do this after the CL system had
> booted:
>
> Start mount
> Start nmbd
> Start Smdb
> Start docker (plex)
>
>
> On 21 Nov. 2017 08:25, "Douglas, William" <william.douglas(a)intel.com> wrote:
>
> You might be seeing an instance where /dev/disk/by-uuid is not fully
> populated by the time your service runs. Having your disk attached to
> an adapter makes this more likely to me.
>
> Maybe try
> https://www.freedesktop.org/software/systemd/man/systemd.mount.html#Timeout…
> but that's just for the mount command so I'm not sure if it will even
> attempt to run the mount or not based on the uuid missing log you had
> but that's the only thing I could find that might work.
>
> On Mon, Nov 20, 2017 at 1:08 PM, Richard Thornton
> <richie.thornton(a)gmail.com> wrote:
>> I should add that the btrfs disks are attached to an LSI pcie adapter.
>>
>> On 21 Nov. 2017 07:56, "Richard Thornton" <richie.thornton(a)gmail.com>
>> wrote:
>>>
>>> My dir:
>>>
>>> root(a)thor ~ # ls /etc/systemd/system/
>>>
>>> multi-user.target.wants nmbd.service smbd.service storage.mount
>>>
>>> On 21 November 2017 at 07:42, Richard Thornton
>>> <richie.thornton(a)gmail.com> wrote:
>>> > dmesg snippet:
>>> >
>>> > [ 30.073488] BTRFS: device fsid a4d1ef9d-e75f-41dd-acac-9aaa79428965
>>> > devid 2 transid 16 /dev/sdb
>>> > [ 30.073911] BTRFS info (device sdb): disk space caching is enabled
>>> > [ 30.073912] BTRFS info (device sdb): has skinny extents
>>> > [ 30.074170] BTRFS warning (device sdb): devid 1 uuid
>>> > fc680ff9-c623-4baf-85d4-c634d2dbc6ef is missing
>>> > [ 30.074171] BTRFS error (device sdb): failed to read the system
>>> > array: -5
>>> > [ 30.101015] BTRFS error (device sdb): open_ctree failed
>>> >
>>> > On 21 November 2017 at 07:21, Richard Thornton
>>> > <richie.thornton(a)gmail.com> wrote:
>>> >> Thanks, the below works with "systemctl start storage.mount" but not
>>> >> when enabled for boot:
>>> >>
>>> >> [Unit]
>>> >> Description=Mount
>>> >> Before=nmbd.service
>>> >>
>>> >> [Mount]
>>> >> What=/dev/disk/by-uuid/a4d1ef9d-e75f-41dd-acac-9aaa79428965
>>> >> Where=/storage
>>> >>
>>> >> [Install]
>>> >> WantedBy=multi-user.target
>>> >>
>>> >> On 21 November 2017 at 06:04, Douglas, William
>>> >> <william.douglas(a)intel.com> wrote:
>>> >>> So I didn't see any issues when I did this. Can you show the
>>> >>> storage.mount file and tell me what is the target that requires
>>> >>> storage.mount?
>>> >>>
>>> >>> On Mon, Nov 20, 2017 at 8:49 AM, Douglas, William
>>> >>> <william.douglas(a)intel.com> wrote:
>>> >>>> I haven't tried this recently so I'll give it a go today and get
>>> >>>> back
>>> >>>> to
>>> >>>> you.
>>> >>>>
>>> >>>>
>>> >>>> On Nov 20, 2017 02:45, "Richard Thornton"
>>> >>>> <richie.thornton(a)gmail.com>
>>> >>>> wrote:
>>> >>>>
>>> >>>> Thanks, I finally got around to setting this up, systemd is loading
>>> >>>> samba on boot and I could mount btrfs with
>>> >>>> "systemctl start storage.mount" after boot.
>>> >>>>
>>> >>>> The issue I'm having is I can't get a btrfs to mount on boot, I
>>> >>>> tried
>>> >>>> systemd and fstab. I think it could be something to do with this:
>>> >>>>
>>> >>>>
>>> >>>>
>>> >>>> https://blogs.gentoo.org/remi/2014/12/05/gentoo-btrfs-arrays-and-systemd-a-…
>>> >>>>
>>> >>>> I spent hours on trying to get this to persistent mount, including
>>> >>>> doing what the above link stated, can someone please help me out?
>>> >>>>
>>> >>>> Thanks.
>>> >>>>
>>> >>>> Richard
>>> >>>>
>>> >>>> On 20 October 2017 at 02:19, Nesius, Robert A
>>> >>>> <robert.a.nesius(a)intel.com>
>>> >>>> wrote:
>>> >>>>> Yep - you got it.
>>> >>>>>
>>> >>>>> Technically systemd supports /etc/fstab for backwards compatibility
>>> >>>>> - it
>>> >>>>> generates a mount unit for each line in the file.
>>> >>>>>
>>> >>>>> -Rob
>>> >>>>>
>>> >>>>> Sent from mobile device.
>>> >>>>>
>>> >>>>> On Oct 18, 2017, at 9:03 PM, Richard Thornton
>>> >>>>> <richie.thornton(a)gmail.com>
>>> >>>>> wrote:
>>> >>>>>
>>> >>>>> Hi,
>>> >>>>>
>>> >>>>> Sorry to bother you all.
>>> >>>>>
>>> >>>>> I like Clear Linux a lot and Arjan has been a great help.
>>> >>>>>
>>> >>>>> Posted to dev@ because I couldn't see a user@
>>> >>>>>
>>> >>>>> I'm going to use Clear Linux for a 24/7 home server, I like the
>>> >>>>> upgrade system (used to use Ubuntu).
>>> >>>>>
>>> >>>>> I'm going to use Docker containers for my applications.
>>> >>>>>
>>> >>>>> I want to setup some persistence, if there is a reboot I would like
>>> >>>>> my
>>> >>>>> storage disks to mount, Samba to start and my Docker containers to
>>> >>>>> run.
>>> >>>>>
>>> >>>>> I think the way to do this is through systemd on CL, if so, I need
>>> >>>>> to
>>> >>>>> create the /etc/systemd structure?
>>> >>>>>
>>> >>>>> Just looking for best practices, advice and any documentation that
>>> >>>>> may
>>> >>>>> be out there.
>>> >>>>>
>>> >>>>> Thanks.
>>> >>>>>
>>> >>>>> Richard
>>> >>>>> _______________________________________________
>>> >>>>> Dev mailing list
>>> >>>>> Dev(a)lists.clearlinux.org
>>> >>>>> https://lists.clearlinux.org/mailman/listinfo/dev
>>> >>>> _______________________________________________
>>> >>>> Dev mailing list
>>> >>>> Dev(a)lists.clearlinux.org
>>> >>>> https://lists.clearlinux.org/mailman/listinfo/dev
>>> >>>>
>>> >>>>
>
>