/etc/hosts?
by David Lee
(cobbler 2.0.11; on RHEL 5.6 server; installing RHEL 5.6 clients)
I'm making good progress with cobbler, and trying to keep things as
simple as possible.
But I've spotted a problem. A freshly installed system (say "client1")
has a hosts file that contains just one active IPv4 line:
127.0.0.1 client1.my.domain client1 localhost.localdomain localhost
Shouldn't that be two lines, something like:
127.0.0.1 localhost.localdomain localhost
10.156.x.y client1.my.domain client1
That is, putting my real hostname and real IP as a separate entry.
This seems so intuitive, and likely to be so commonly wanted, that I
presume there must be some hook in cobbler that I've missed. (Or do I
have to write some sort of post-install snippet to manipulate /etc/hosts
to achieve this?)
--
: David Lee
: ECMWF (Data Handling System)
: Shinfield Park
: Reading RG2 9AX
: Berkshire
:
: tel: +44-118-9499 362
: email: david.lee(a)ecmwf.int
12 years, 8 months
Debian/Ubuntu basics
by petermity
I want to load the Ubuntu 10.04 desktop CD contents into Cobbler
Server so I can network-boot a PC from it and load it without
interaction, like I do with CentOS and Fedora in Cobbler V2.0.11.
I don't need updates or syncing of repos or anything. I've
looked at wiki, bugs, forums, and am fairly confused.
For CentOS/Fedora I used "import" but it doesn't work with
--breed=ubuntu or debian. Some docs say to use "distro add" but
that doesn't let you specify a --path, so how do I get the CD
content into Cobbler with that? I thought maybe "repo add" would
work but it doesn't seem to, it creates a repo entry but never
copies the content of the CD.
12 years, 8 months
list-harddrives macro
by Greg Chavez
With great success, I've used variations on the partition_select
snippet to create some very intelligent partitioning for different
profiles. However, I occasionally come across a scenario whereby my
Dell servers end up with the partitions reversed. For example, most of
our PERC RAID cards are setup like this:
Virtual Disk #0: 2 73G drives, RAID 0 (OS disk)
Virtual Disk #1: 6 300G drives, RAID 5 (data disk)
Generally, VD-0 gets read by anaconda as sda, and VD-1 as sdb. And in
fact, even on the messed up servers this works:
Disk /dev/sda: 72.7 GB, 72746008576 bytes
Disk /dev/sdb: 1497.2 GB, 1497198755840 bytes
But somehow, despite this snippet code (my simplest scenario):
=============================
set \$(list-harddrives)
let numd=\$#/2
d1=\$1
d2=\$3
cat << EOF > /tmp/partinfo
part /boot --fstype ext3 --size 200 --ondisk=\$d1 --asprimary
part swap --recommended --ondisk=\$d1 --asprimary
part pv.root --size=1024 --grow --ondisk=\$d1
part pv.srv --size=1024 --grow --ondisk=\$d2
volgroup vg00 pv.root
volgroup vg01 pv.srv
logvol / --fstype ext3 --size=1024 --vgname=vg00 --name=root --grow
logvol /srv --fstype ext3 --size=1024 --vgname=vg01 --name=srv --grow
EOF
=============================
The pv.root gets assigned to volume group 00, and pv.srv gets assigned
to volume group 01 and I get this:
[root@io-kvm-16 ~]# pvs
PV VG Fmt Attr PSize PFree
/dev/sda2 vg01 lvm2 a- 67.53g 0
/dev/sdb2 vg00 lvm2 a- 1.34t 0
AND
[root@io-kvm-16 ~]# df -h / /srv
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/vg00-root
1.4T 1.8G 1.3T 1% /
/dev/mapper/vg01-srv 67G 180M 63G 1% /srv
Argh! My suspicion is that the list-harddrives macros is spitting out
the devices names in reverse, such that $d1 gets sdb and $d2 gets sda.
Does anybody know where the list-harddrives macro comes from? Is it
an anaconda thing? How does it work? I realize that this is not
really a Cobbler issue, but maybe some of you who use this snippet
know more about my situation in general and this macro in particular?
Any help you guys can give me would be much appreciated, even if it's
just pointing me to another mailing list.
--
--Greg Chavez
--
12 years, 8 months
unable to retrieve stage2.img after cobbler-web install
by Bruijnes, A. (Arno)
Hello,
I want to add systems with cobbler via the webgui. I have follow this
link :https://fedorahosted.org/cobbler/wiki/CobblerWebInterface.
When i add a system and will install RHEL, i have this error:
Unable to retrieve
http://satellite server//ks/dist/ks distribution/images/stage2.img
When i remove the cobbler-web package and add a system with cobbler
command line it is working.
The following conf file is added after install cobbler-web and i think
there is a problem
# cat cobbler_web.conf.rpmsave
# This configuration file enables the cobbler web
# interface (django version)
# Do not log the requests generated from the event notification system
SetEnvIf Request_URI ".*/op/events/user/.*" dontlog
# Log only what remains
CustomLog logs/access_log combined env=!dontlog
<Location "/cobbler_web">
SetHandler python-program
PythonHandler django.core.handlers.modpython
SetEnv DJANGO_SETTINGS_MODULE settings
# PythonOption django.root /cobbler_web
PythonDebug On
PythonPath "['/usr/share/cobbler/web/'] + sys.path"
AuthBasicAuthoritative Off
AuthType basic
AuthName "Cobbler"
Require valid-user
PythonAuthenHandler cobbler_web.views
</Location>
Met vriendelijke groet,
Arno Bruijnes.
-------------------------Disclaimer-------------------------------
De informatie verzonden met dit e-mailbericht (en bijlagen) is uitsluitend bestemd voor de geadresseerde(n) en zij die van de geadresseerde(n) toestemming kregen dit bericht te lezen. Gebruik door anderen dan geadresseerde(n) is verboden. De informatie in dit e-mailbericht (en bijlagen) kan vertrouwelijk van aard zijn en kan binnen het bereik vallen van een geheimhoudingsplicht en een verschoningsrecht.
-------------------------------------------------------------------
12 years, 8 months
Corrupt/missing RPM's - how to repair ?
by Dan White
Trying to PXE-boot/install systems, I hit a virtual brick wall due to a missing or corrupt RPM
Lacking any clue, I deleted all references to the offending package from my cobbler-server, resync-ed and I am now running a cobbler reposync in the hopes that the offender will get replaced.
Did I do something horribly wrong or what ?
“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, 8 months
'change' triggers - what are they for and how are they called?
by Stuart Sears
Hi, list!
So, after replying to Paul's question about ksmeta variables, I decided
to experiment with creating a trigger to render a custom per_system
snippet with ks metadata args, for reporting if nothing else.
This page has been helpful (and should be linked from the cobbler wiki
IMO. So I have :):
http://www.ithiriel.com/content/2010/03/29/writing-install-triggers-cobbler
So I tried something similar to the above as a pure python trigger, but
this eventually, after much facepalming, exposed that 'change' triggers
appear to be called with no args whatsoever, so I am unable to determine
what precisely has been changed.
Why would this be? By design?
Am I completely misunderstanding them?
If I wanted, for example to record the changes made to ksmeta etc for
profiles or systems (and possibly other objects), can I do this?
has anyone successfully used these triggers?
(Puzzled) Regards
Stuart Sears
12 years, 8 months
Cobbler and Active Directory DNS
by McEvoy, James
Has anyone used cobbler to create/maintain DNS records for an Active Directory DNS server? We are developing a Linux NAS server that will run both NFS and CIFS protocols to serve files so it will be easier to use Active Directory authentication and DNS in this setting... Sharing any experiences or ideas on how to make this work will be appreciated.
--jim
12 years, 8 months
Setting netmask/subnet with the xmlrpc in a python script
by Thomas Ewert
I am developping a webinterface for a customer to set up new systems
in cobbler. I am using the xmlrpc.
using it in that way:
.......
obj = server.new_system(token)
server.modify_system(obj,"name",bezeichnung, token)
server.modify_system(obj,"profile",profil, token)
server.modify_system(obj,"mgmt_classes",[ application ], token)
server.modify_system(obj,"comment", comment , token)
.............................
Now I have the problem to set the netmask for a interface (eth0, eth1...)
Can anyone help please??
For Info: Ilearned python by my own about 3 weeks ago and also cobbler.
BR from Munich
Tom
--
Thomas Ewert
Birkenweg 11
82291 Mammendof
0178 8855 733
Thomas.Ewert(a)googlemail.com
12 years, 8 months
Directories for Advanced Snippets ?
by Dan White
In the wiki at <https://fedorahosted.org/cobbler/wiki/
KickstartSnippets> it talks about a tree of snippets --
/var/lib/cobbler/snippets/per_system/$snippet_name/$system_name
/var/lib/cobbler/snippets/per_profile/$snippet_name/$profile_name
/var/lib/cobbler/snippets/per_distro/$snippet_name/$distro_name
/var/lib/cobbler/snippets/$snippet_name
Does something in cobbler generate that tree for you or do you have to
build it by hand ?
12 years, 8 months
Trouble Adding Distros: Internal Error Unknown Distro Name
by Pete Eby
Howdy,
Though I've installed Cobbler several times in different environments,
usually with little issue, I've run into a snag which I'm having
trouble working out.
I'm running cobbler-2.0.11-2.el6.noarch and (cobbler-web) from EPEL on
RHEL 6.1 and I can't seem to be able to add a distro.
When I run:
cobbler distro add --breed=redhat --name=CentOS-5.6-32bit
--kernel=/var/lib/tftpboot/distros/CentOS5.6-32bit/vmlinuz
initrd=/var/lib/tftpboot/distros/CentOS5.6-32bit/initrd.img
No error is sent to std out, but then when I do a cobbler distro list,
no results are returned.
Looking at the log I see:
Tue Jul 26 09:22:50 2011 - INFO | REMOTE new_item(distro); user(<DIRECT>)
raise CX("internal error, unknown %s name %s" % (what,name))
File "/usr/lib/python2.6/site-packages/cobbler/remote.py", line
633, in get_item_handle
handle = self.get_item_handle(object_type, object_name)
File "/usr/lib/python2.6/site-packages/cobbler/remote.py", line 798,
in xapi_object_edit
Tue Jul 26 09:22:50 2011 - INFO | Exception Info:
Tue Jul 26 09:22:50 2011 - INFO | Exception value: 'internal error,
unknown distro name CentOS-5.6-32bit'
Tue Jul 26 09:22:50 2011 - INFO | Exception occured: <class
'cobbler.cexceptions.CX'>
Tue Jul 26 09:22:50 2011 - DEBUG | get_item; ['distro', 'CentOS-5.6-32bit']
Tue Jul 26 09:22:50 2011 - DEBUG | REMOTE CLI Authorized; user(?)
I've tried some variations, but get the above in all cases.
Any suggestion would by much appreciated.
Thank you,
Pete
12 years, 8 months
