Some questions regarding extension of cobbler suse snippets/templates
by Christian Horn
Hi cobbler-devel,
there are some suse sles related things that could make sense to
be included directly in cobbler.
These pieces are currently here:
https//fluxcoil.net/doku.php/cobbler/sles .
Uwe, I think the existing suse pieces are from you, so you
might be able to point out some things I am missing.
Here are my questions, please comment if things make sense
to get included and in which way:
- Currently cobbler comes with autoyast_sample.xml, I think of additionally
shipping a autoyast_complex.xml, with these things included:
- snippet which sets up kdump
- a construct which allows sles-rootpw to be set if specified
as ksmeta $rootpw
- code which allows selection of one of 3 package selections
which are selected via a ksmeta, minimal/tiny/big
- post snippet: disabling of some services
- post snippet: this deploys a snippet with pure shellcode
(which works also for rhel and cleans up /etc/grub.conf),
and a wrappersnippet. The wrappersnippet makes it possible
reuse snippets written for rhel (containing shellcode) also for
sles (in wrapping it in the xml code that autoyast expects).
Which of these look useful for inclusion? Should documentation happen
inline as comments or in wiki?
- While the above works here for sles11, with some functions dropped I have this
also here for sles9 and 10. These are separate template files, and separate
interface snippets for these other distros. Is this of interest to be included?
- currently, when the networking.xml snippet is included and the hostname
does not contain a hostname it bails out with an error message.
In the code here this is checked for using
#if re.match('.*\..*', $hostname)
- the current networking.xml is not adapted to use $network instead
of $subnet .
The scripts here cope with both now, when changing this in git then
only the $netmask has to be used I guess?
- current cobbler git head uses the SuSE subdirectory, I am currently only
testing with cobbler 2.2 in fedora updates-testing.
- currently, autoyast.xml is not directly usable as cobbler template.
It is xml valid, but the <!-- comments --> at the beginning seems to be the
problem. So users have to remove this bevore using.
I would suggest this file no longer be xml valid, instead using here
## these
## comments
which are filtered out by cobbler later.
Opinions on the points above welcome.
cheers, Christian
12 years, 2 months
Re: Side issue: Subversion support ? Re: managing cobbler with puppet / augeas
by Dan White
If I was in control of the Subversion server, I would be happy to try Git instead, but all I control is Cobbler, so I would like to see if I can modify it.
Am I correct in that the mods should be able to live in cobbler/modules/scm_track.py ?
Where I need help is that I'm an old C/C++ coder who is used to compiling his code.
I am still learning Pythin, and loving it.
What do I need to do to "compile" my changes to the Python source into the system to then test it ?
“Sometimes I think the surest sign that intelligent life exists elsewhere in the universe is that none of it has tried to contact us.”
Bill Waterson (Calvin & Hobbes)
----- Michael DeHaan <michael.dehaan(a)gmail.com> wrote:
>
>
> On Tuesday, January 31, 2012 at 8:57 AM, Dan White wrote:
>
> > Hey Mike !
> > Replying directly to you rather than look like a thread-jacker :)
> >
> > I just tossed Issue #59 into the pot -- Add Subversion support to cobbler/modules/scm_track.py
> >
> > I am willing to do it, but I need some info about how to proceed within Cobbler's environment.
>
> The best tip for SVN users is to switch to git ASAP :)
>
> That being said, let us know what you need to know and we'll help as we can. I recommend the cobbler-devel list.
>
>
12 years, 2 months
Self introduction and Power management improvements
by Arnaud Quette
Dear Cobbler fellows,
First, to introduce myself, I'm Arnaud Quette.
I'm mainly in charge of the NUT - Network UPS Tools [1] project.
NUT is the main project for power devices (UPS, PDU, PSU, ... ) management
and protection, and provides the broadest list of supported devices in the
field [2].
I'm also working on several infrastructure and cloud power management
related topics.
As such, I've been looking at Cobbler in 2 ways:
1) improving existing power management (Ie, fence-agents integration)
2) proposing a design (and latter implementing) for power protection.
For point (1), I've already made a few fixes, with a stagging pull request
on Github [3]:
- actually make a combo list for "Type", with a sorted list of scripts
(since it's a finite list, depending on the available power templates),
- Reword "Type" tooltip, for more clarity,
- Rename from "Power" to "Power management", to prepare for the upcoming
"Power protection" section (latter RFC),
- Remove "Power" from the various fields (Type, Address, Username,
Password, ID) to avoid redundancy.
I'm also working on creating a "fence_nut" agent.
Once submitted upstream to Marek (Grac) and Fabio (M. Di Nitto), I'll also
submit the Cobbler template through a pull request...
For point (2), I'm still working on a design to submit you, but have a few
questions to ask you you before.
- the general idea is to make it easy to deploy power protection on the
provisioned servers, using the NUT packages,
- this implies some configuration, to point at the protecting and feeding
devices (UPS at least, but also PDU and servers power supply units),
- I'm intending to create a new "Power protection" section, still in the
system page, to allow for setting up these values,
I've made good progress on this part.
- I'm working with inventory systems (OCS Inventory NG and Fusion
Inventory) to discover power devices (through USB, SNMP and IPMI).
The aim is then to be able to retrieve inventory data into cobbler (through
web service interfaces), and provide users with the list of physical
servers and power devices to help filling the various forms .
As a complement, NUT also provides a tool (nut-scanner, which is used by
inventory systems), to discover NUT supported devices.
It is currently available as a command line, but Python and Perl wrapper
are underway.
This can be a fallback, in case inventory systems are not available.
Now, the questions:
- do you have any input or comment on point (1) and (2)?
- once everything is setup in the UI, how can we enable installation of NUT
packages and push the configuration done by the user?
I'll get back to you soon with more details on point (2).
cheers,
Arnaud
--
[1] http://www.networkupstools.org/
[2] http://www.networkupstools.org/stable-hcl.html
[3] https://github.com/cobbler/cobbler/pull/46
--
Linux / Unix Expert R&D - Eaton - http://powerquality.eaton.com
Network UPS Tools (NUT) Project Leader - http://www.networkupstools.org/
Debian Developer - http://www.debian.org
Free Software Developer - http://arnaud.quette.free.fr/
12 years, 3 months
Anyone got some cheese to go with this whine ? :) Build-ISO problems in search of a solution
by Dan White
First, I have something to contribute before I start whining --
I found a workflow for RHEL 5 to get a cobbler generated build ISO onto a USB flash drive
Install LiveCD-Tools ( yum install livecd-tools )
Partition the USB thumb with : fdisk /dev/sdb
Delete all partitions and then make one that covers the whold device.
Format it with: mkfs.ext3 /dev/sdb1
I made the ISO with: cobbler buildiso --systems=mynewservername.foo.bar
mynewservername.foo.bar having been previously defined in cobbler with profile, etc.
Then the magic: /usr/bin/livecd-iso-to-disk /path/to/cobbler/build-iso/generated.iso /dev/sdb1
This boots up my machine and gives me the menu to select from.
I selected the system name (mynewservername.foo.bar) and started it.
It asks me to pick the NIC to run from.
These servers have 4 built-in ethernet ports.
I found that a bit surprising as I defined only one port (eth0) in the cobbler system.
I would hope it would take that and run with it.
Had I defined more than one, I could understand the query.
After that, the kickstart ran as expected, up to and including the reboot.
During reboot, I removed the USB thumb, expecting its job finished.
When the machine comes back up, it fails to load the operating system.
I did not record the exact screen messages, but it indicated (to me) a corrupted OS load.
I found a cause for this, but not a solution: The contents of the USB flash drive had been overwritten during the OS installation process ! I have write-protectable USB flash drives from RiData/RiTek (product name: Twister), but did not have one handy to try.
The ISO I put onto this USB thumb (at least twice) totally failed to load my server as expected.
On a happy note, the same ISO burned onto a CD-ROM and re-attempted with a CD drive ran like a champ.
I do not have the luxury to do a lot of pounding on this, but I will try it once more with a write-protectable USB flash drive to see what happens.
At this point, I'd advise folks to waste the CD's than fuss with a USB thumb.
I have an alternate workflow to suggest for this, but I will post it separately
Finally, would you consider this a "bug" ? If so, I will report it that way as well.
Thanks for listening. The cheese was great(ed).
“Sometimes I think the surest sign that intelligent life exists elsewhere in the universe is that none of it has tried to contact us.”
Bill Waterson (Calvin & Hobbes)
12 years, 3 months
Re: Anyone got some cheese to go with this whine ? :) Build-ISOproblems in search of a solution
by Simon Woolsgrove
As Stuart has, you can also select the the interface is a management interface in the Web UI or use --management from the CLI when you add or edit an interface, this is passed to buildiso and adds the ksdevice=xxxx param and ip details if using static to your boot line.
Cheers,
Simon
On Mon 16/01/12 11:27 AM , Stuart Sears <stuart(a)sjsears.com> wrote:
> Hi Dan,
>
> responses below...
>
> On 13/01/12 21:20, Dan White wrote:
> [...]
> > It asks me to pick the NIC to run from. These servers have 4 built-in
> > ethernet ports. I found that a bit surprising as I defined only one
> > port (eth0) in the cobbler system. I would hope it would take that
> > and run with it. Had I defined more than one, I could understand the
> > query.
>
> You need a --ksdevice=eth0 argument to the kernel
> add it to your 'kernel-options' for the system record.
>
> Cobbler doesn't actually do the installation, it just creates a
> kickstart file, which is processed by anaconda. Anaconda has no idea
> which of the 4 devices you have you would like to install through.
> This has been an 'interesting' point of discussion for a long time now,
> as I recall.
>
> > After that, the kickstart ran as expected, up to and including the
> > reboot.
> >
> > During reboot, I removed the USB thumb, expecting its job finished.
> > When the machine comes back up, it fails to load the operating
> > system. I did not record the exact screen messages, but it indicated
> > (to me) a corrupted OS load.
> >
> > I found a cause for this, but not a solution: The contents of the
> > USB flash drive had been overwritten during the OS installation
> > process ! I have write-protectable USB flash drives from
> > RiData/RiTek (product name: Twister), but did not have one handy to
> > try.
> >
> > The ISO I put onto this USB thumb (at least twice) totally failed to
> > load my server as expected.
>
> Does it boot with the usb key in afterwards? (see my bootloader comment
> below)
>
> > On a happy note, the same ISO burned onto a CD-ROM and re-attempted
> > with a CD drive ran like a champ. I do not have the luxury to do a
> > lot of pounding on this, but I will try it once more with a
> > write-protectable USB flash drive to see what happens.
>
> This is a kickstart default, not a cobbler problem.
> check the rendered kickstart for a line like this:
>
> clearpart --all
>
> This is probably what is doing that.
> It may well be installing the bootloader on the usb stick for all I
> know. If you know the names of the internal disks then you can put
>
> clearpart --all --drives=sda (for example)
>
> Alternatively anaconda/kickstart has an 'ignoredisk' option you can use
> to avoid this.
>
> ignoredisk --onlyuse sda
>
> (or possibly /dev/sda. Can't remember right now)
>
> > At this point, I'd advise folks to waste the CD's than fuss with a
> > USB thumb.
> >
> > I have an alternate workflow to suggest for this, but I will post it
> > separately
> >
> > Finally, would you consider this a "bug" ? If so, I will report it
> > that way as well.
>
> No. not in cobbler at any rate.
>
> Stuart
> --
> Stuart Sears RHCA etc.
> "It's today!" said Piglet.
> "My favourite day," said Pooh.
> _______________________________________________
> cobbler-devel mailing list
> cobbler-devel(a)lists.fedorahosted.org
> https://fedorahosted.org/mailman/listinfo/cobbler-devel [1]
>
>
>
> Links:
> ------
> [1]
> http://woolsgrove.com/parse.php?redirect=https://fedorahosted.org/mailman/l
> istinfo/cobbler-devel
>
12 years, 3 months
Re: cobbler bootiso ontimeout option
by Simon Woolsgrove
Thats not an option as the action_buildiso code is inserting syslinux menu commands, also I only need this for virtual for physical I would want default Local in place e.g. hence supplying an option to buildiso would be nice.
I think this is probably just a use case for this particular env, I will take a look at a patch and see if it makes sense.
Cheers,
Simon
On Tue 03/01/12 12:54 PM , Michael DeHaan <michael.dehaan(a)gmail.com> wrote:
> "Right you can change the bootiso template but the ontimeout value needs to
> match the System or Profile name your requesting,"
> If you know what it's going to timeout to, you should just set the IP/MAC
> so it boots directly and not use the menu, otherwise you won't know what
> you are supposed to default to anyway.
> -- Michael
>
> On Tuesday, January 3, 2012 at 6:07 AM, Simon Woolsgrove wrote:
>
> Right you can change the bootiso template but the ontimeout value needs to
> match the System or Profile name your requesting, in this use case its
> always a system since were creating single system iso's passing the ip
> details on the boot line, call it a very poor replacement for dhcp. We can
> not use the same default name over and over, the buildiso template does not
> support replacing variables that I am aware? like say in the pxe template
> but it should not be too much to add this.
>
> Cheers,Simon
>
> On Fri 30/12/11 4:25 PM , Michael DeHaan wrote:
> Ah, you said ontimeout? Try editing the templates in /etc/cobbler…
> the"local" part is in that template, not the cobbler source, so you can
> changeit.
> -- Michael
> On Friday, December 30, 2011 at 11:20 AM, Michael DeHaan wrote: Thisfeature
> already exists. Create a system named "default":cobbler system add
> --name=default --profile=fooPXE also works by IP, so it at least used to
> work that you could create asystem ip in CIDR notation to get the same
> thing for just a part of yournetwork:cobbler system add
> --name=thisNetworkHasADefault --ip=255.255.255.0/16 --profile=fooThe latter
> may o --Michael
> -- Michael
> On Friday, December 30, 2011 at 7:27 AM, Jörgen Maas wrote: On Fri, Dec30,
> 2011 at 1:10 PM, wrote:
> Have a use case that maybe unique to us as were not unfortunately
> usingDHCPor all our vlans. We automate on demand VMware guests by using
> aworkflowengine to; create the guest images, cobbler system record, request
> cobblerbuildiso, attach iso and boot guest, detach iso. Our issue is
> thenthat on the isolinux timeout the default boot is Local which is fine
> for99% ofcases but not this as the guest build fails. Currently we
> haveanother stepto mangle the cd athough it occurs the better solution
> maybe toadd anotheroption say ontimeout so that can be supplied via
> api/cli (notsure it makessense in the webui)?Hi Simon, so you want to be
> able to control the value of the ONTIMEOUT. So that you can select a system
> to install without any further intervention?--
> Grtz,JörgenMaas_______________________________________________cobbler-deve
> lmailinglistcobbler-devel@lists.fedorahosted.orghttps://fedorahosted.org/mai
> lman/li[1]stinfo/cobbler-devel[1]
> _______________________________________________ cobbler-devel mailing
> listcobbler-devel@lists.fedorahosted.orghttps://fedorahosted.org/mailman/li
> stinfo/cobbler-devel[2] [2]
>
> Links:------[1] https://fedorahosted.org/mailman/listinfo/cobbler-devel [3]
> [2]
>
> Links:
> ------
> [1] http://fedorahosted.org/mailman/li
> [2] https://fedorahosted.org/mailman/listinfo/cobbler-devel
> [3] https://fedorahosted.org/mailman/listinfo/cobbler-devel
> [4] https://fedorahosted.org/mailman/listinfo/cobbler-devel
>
>
12 years, 3 months
Re: cobbler bootiso ontimeout option
by Simon Woolsgrove
Right you can change the bootiso template but the ontimeout value needs to match the System or Profile name your requesting, in this use case its always a system since were creating single system iso's passing the ip details on the boot line, call it a very poor replacement for dhcp. We can not use the same default name over and over, the buildiso template does not support replacing variables that I am aware? like say in the pxe template but it should not be too much to add this.
Cheers,
Simon
On Fri 30/12/11 4:25 PM , Michael DeHaan <michael.dehaan(a)gmail.com> wrote:
> Ah, you said ontimeout? Try editing the templates in /etc/cobbler… the
> "local" part is in that template, not the cobbler source, so you can change
> it.
>
> -- Michael
>
> On Friday, December 30, 2011 at 11:20 AM, Michael DeHaan wrote: This
> feature already exists. Create a system named "default":
> cobbler system add --name=default --profile=foo
> PXE also works by IP, so it at least used to work that you could create a
> system ip in CIDR notation to get the same thing for just a part of your
> network:
> cobbler system add --name=thisNetworkHasADefault --ip=255.255.255.0/16
> --profile=foo
> The latter may o
> --Michael
>
> -- Michael
>
> On Friday, December 30, 2011 at 7:27 AM, Jörgen Maas wrote: On Fri, Dec
> 30, 2011 at 1:10 PM, wrote:
>
> Have a use case that maybe unique to us as were not unfortunately using
> DHCPor all our vlans. We automate on demand VMware guests by using a
> workflowengine to; create the guest images, cobbler system record, request
> cobblerbuildiso, attach iso and boot guest, detach iso. Our issue is then
> that on the isolinux timeout the default boot is Local which is fine for
> 99% ofcases but not this as the guest build fails. Currently we have
> another stepto mangle the cd athough it occurs the better solution maybe to
> add anotheroption say ontimeout so that can be supplied via api/cli (not
> sure it makessense in the webui)?
> Hi Simon, so you want to be able to control the value of the ONTIMEOUT. So
> that you can select a system to install without any further intervention?
> -- Grtz,Jörgen
> Maas_______________________________________________cobbler-devel mailing
> listcobbler-devel@lists.fedorahosted.orghttps://fedorahosted.org/mailman/li
> stinfo/cobbler-devel[1]
>
> _______________________________________________
> cobbler-devel mailing list
> cobbler-devel(a)lists.fedorahosted.org
> https://fedorahosted.org/mailman/listinfo/cobbler-devel [2]
>
>
>
> Links:
> ------
> [1] https://fedorahosted.org/mailman/listinfo/cobbler-devel
> [2]
> http://woolsgrove.com/parse.php?redirect=https://fedorahosted.org/mailman/l
> istinfo/cobbler-devel
>
12 years, 3 months