Need help ... Urgently ...
by Yogesh Mulay
I have the Cobbler setup ready & esxi5 installed. Is there a way to parse
esxi commands in the host. Like enabling ssh running esxcli. Adding VMKs.
Changing the managament IP address & its Vlan
I need some tips regarding pre & post installation basically
Any help would be highly appreciated. ...
Regards
10 years
<ProtocolError for localhost:25151/cobbler_api: 404 Not Found>
by John Paget Bourke
Hi,
When I run
koan --list=profiles -s localhost
I get the following error
root@Config cobbler]# koan --list=profiles -s localhost
- looking for Cobbler at http://localhost:25151/cobbler_api
Traceback (most recent call last):
File "/usr/lib/python2.6/site-packages/koan/utils.py", line 542, in
__try_connect
xmlrpc_server.ping()
File "/usr/lib64/python2.6/xmlrpclib.py", line 1199, in __call__
return self.__send(self.__name, args)
File "/usr/lib64/python2.6/xmlrpclib.py", line 1489, in __request
verbose=self.__verbose
File "/usr/lib64/python2.6/xmlrpclib.py", line 1243, in request
headers
ProtocolError: <ProtocolError for localhost:25151/cobbler_api: 404 Not
Found>
If I try to see if the server is running and listening, the server is ok
[root@Config cobbler]# netstat -an | grep 25151
tcp 0 0 127.0.0.1:25151 0.0.0.0:*
LISTEN
tcp 0 0 127.0.0.1:25151 127.0.0.1:57483
TIME_WAIT
But returns nothing .
[root@Config cobbler]# curl http://localhost:25151/cobbler_api
curl: (52) Empty reply from server
I am using the web GUI and it works ok, no problems.
When I try to use koan or curl and check the logs
[root@Config cobbler]# curl http://localhost:25151/cobbler_api
curl: (52) Empty reply from server
[root@Config cobbler]# date
Sun Oct 30 05:30:55 UTC 2011
[root@Config cobbler]# ls -l
total 1792
drwxr-xr-x. 2 root root 4096 Oct 6 02:24 anamon
-rw-r--r--. 1 root root 1806499 Oct 30 04:50 cobbler.log
-rw-r--r-- 1 root root 73 Oct 29 00:24 install.log
drwxr-xr-x. 2 root root 4096 Oct 6 02:24 kicklog
drwxr-xr-x. 2 root root 4096 Oct 6 02:24 syslog
drwxr-xr-x. 2 root root 4096 Oct 30 04:40 tasks
[root@Config cobbler]# cd /etc/httpd/logs/
[root@Config logs]# ls -l
total 3172
-rw-r--r-- 1 root root 10904 Oct 30 04:45 access_log
-rw-r--r--. 1 root root 789779 Oct 30 02:43 access_log-20111030
-rw-r--r-- 1 root root 32638 Oct 30 05:20 error_log
-rw-r--r--. 1 root root 2403267 Oct 30 03:06 error_log-20111030
The logs do not seem to be affected.
The versions I am running are
[root@Config cobbler]# rpm -qa | grep cobbler
cobbler-2.2.1-1.el6.noarch
cobbler-web-2.2.1-1.el6.noarch
[root@Config cobbler]# rpm -qa | grep koan
koan-2.2.1-1.el6.noarch
In the modules.conf I have
[authentication]
#module = authn_denyall
module = authn_configfile
[authorization]
module = authz_allowall
So I am a bit lost. Any thoughts ?
Thanks
John
10 years, 2 months
Should I create a Repo to mirror my Distro?
by T Dickinson
I would like my RHEL/CentOS installed servers (provisioned via cobbler) to refer back to the cobbler server for all yum updates/installs.
I understand that is easy for mirrored repos, but how do I configure cobbler so that the distro is used (after the kickstart is over) ?
Im using 2.4 should my installed server's yum repos point at the cobbler server for the base repos? I'm using CentOS-5.9 and /etc/yum.repos.d/CentOS-Base.repo does not point to my server hence any post install yum calls go to internet. Should I create a Repo for the Distro I've installed… this post hints at my dilema http://www.redhat.com/archives/et-mgmt-tools/2007-September/msg00128.html
10 years, 6 months
Rendering oddness in 2.4.0
by Corey Kovacs
I recently bestowed the virtues of cobbler unto a colleague who has, like
many, decided it's the best thing since sliced bread. Good so far...
After he had been working with it for a while, he experienced a problem
which was keeping his lower per_* snippets from overriding the top level
counterparts.
For example, he has
snippets/post_install
and also
snippets/per_system/post_install/mysystem01.
In both files he had (unbeknownst before the override) unescaped "$" chars.
When the per_system post_install snippet was not applied, everything
rendered correctly. However when a specific host was specified in order to
get the overriding snippet, the rendering failed.
So, to recap, two copies of the same messed up snippet were in place. The
top level one rendered correctly, the per_system one did not as cheetah
choked on the unescaped "$" signs.
Is this a know bug/feature?
Corey
10 years, 6 months
Cobbler 2.4 with LDAP(AD) auth
by Danilo Chilene
Hello,
I'm trying setup Cobbler with LDAP(AD), below the configuration of
/etc/cobbler/modules.conf and /etc/cobbler/setup:
modules.conf:
[authentication]
module = authn_ldap
[authorization]
module = authz_allowall
setup:
ldap_server: "192.168.0.1"
ldap_base_dn: "dc=domain,dc=com"
ldap_port: 389
ldap_tls: 0
ldap_anonymous_bind: 0
ldap_search_bind_dn: 'CN=Admin,OU=Users,dc=domain,dc=com'
ldap_search_passwd: strangepassword'
ldap_search_prefix: 'sAMAccountName'
ldap_tls_cacertfile: ''
ldap_tls_keyfile: ''
ldap_tls_certfile: ''
Cobbler Version:
cobbler-2.4.0-1.el6.noarch
cobbler-web-2.4.0-1.el6.noarch
Error:
Mon Sep 30 14:47:05 2013 - INFO | Exception occured: <class
'ldap.FILTER_ERROR'>
Mon Sep 30 14:47:05 2013 - INFO | Exception value: {'desc': 'Bad search
filter'}
Mon Sep 30 14:47:05 2013 - INFO | Exception Info:
File "/usr/lib/python2.6/site-packages/cobbler/remote.py", line 2049, in
_dispatch
return method_handle(*params)
File "/usr/lib/python2.6/site-packages/cobbler/remote.py", line 1840, in
login
if self.__validate_user(login_user,login_password):
File "/usr/lib/python2.6/site-packages/cobbler/remote.py", line 1737, in
__validate_user
return self.api.authenticate(input_user,input_password)
File "/usr/lib/python2.6/site-packages/cobbler/api.py", line 1012, in
authenticate
rc = self.authn.authenticate(self,user,password)
File "/usr/lib/python2.6/site-packages/cobbler/modules/authn_ldap.py",
line 126, in authenticate
result = dir.search_s(basedn, ldap.SCOPE_SUBTREE, filter, [])
File "/usr/lib64/python2.6/site-packages/ldap/ldapobject.py", line 516,
in search_s
return
self.search_ext_s(base,scope,filterstr,attrlist,attrsonly,None,None,timeout=self.timeout)
File "/usr/lib64/python2.6/site-packages/ldap/ldapobject.py", line 509,
in search_ext_s
msgid =
self.search_ext(base,scope,filterstr,attrlist,attrsonly,serverctrls,clientctrls,timeout,sizelimit)
File "/usr/lib64/python2.6/site-packages/ldap/ldapobject.py", line 505,
in search_ext
timeout,sizelimit,
File "/usr/lib64/python2.6/site-packages/ldap/ldapobject.py", line 96,
in _ldap_call
result = func(*args,**kwargs)
I tried different search prefix or even disabled it and no success.
Any ideas on this?
Best Regards,
Danilo F. Chilene
10 years, 7 months
How can I remove parent from a profile?!
by Jakov Sosic
I have a profile that has a parent profile set. I want to remove the
parent profile and set up a distro, preferably without removing the
whole profile...
When I try to remove the parent, i get the following error:
# cobbler profile edit --name=CentOS-6.4-x86_64-test --parent=""
repos
# echo $?
1
Does this sound like a bug?
--
Jakov Sosic
www.srce.unizg.hr
10 years, 7 months
pumpSetupInterface failed create route - 1 Operation not permitted
by Petro
# First interface. VMs get eth0. Physical machines it's usually bond0
interface0='eth0'
ip0='10.67.32.88'
subnet0='255.255.255.240'
gateway0='10.67.32.81'
# Set this to install to a boot lun
# UNSET IT FOR A VM.
#wwid='088d' # formatting note: No colon, lower case.
# Second interface--usually the manglement interface. VMs get eth1.
Physical machines it's usually em2?
interface1='eth1'
ip1='10.67.90.142'
subnet1='255.255.252.0'
# gateway1='10.67.88.1'
This results in network lines that look like:
(sorry about the <thing>.org, but I'm a contractor here and yeah, you know)
network --bootproto=static --ip=10.67.90.142 --netmask=255.255.252.0
--gateway=10.67.32.81 --nameserver=10.65.32.5
--hostname=onpcrfapp01.<thing>.org --device=eth1 --onboot=on
network --bootproto=static --ip=10.67.32.88 --netmask=255.255.255.240
--gateway=10.67.32.81 --nameserver=10.65.32.5
--hostname=onpcrfapp01.<thing>.org --device=eth0 --onboot=on
During the kickstart I get:
22:26:13 INFO : transfering http://server/path to a fd
22:26:13 INFO : setting up kickstart
22:26:13 INFO : kickstartFromURL
22:26:13 INFO : results of url ks, url http:etc. etc. etc.
<load modules>
22.26.15 INFO : need to set up networking
22.26.15 INFO : going to pick interface
22.26.15 INFO : going to do getNetConfig
<etc>
22.26.15 INFO : doing kickstart... setting it up
22.26.15 INFO : result of pumpSetupInterface failed create route - 1
Operation not permitted
oh, and alt-<f4> says:
<6>e1000: eth0 NIC Linuk is up (etc.)
FC stuff...
<6>e1000: eth1 NIC Link is up (etc.)
Google lead me here:
https://access.redhat.com/site/solutions/31028 which asserted that fixing
the network line would solve my problems.
This setup works fine for RHEL6. Unfortunately the product we need to run
is stuck in the sorta cloudy ages and isn't supported on 6, hence trying on
5.
I mangled a bit with cobbler/snippts/network_config to convice cobbler not
put the gateway in eth1, which gave:
network --bootproto=static --ip=10.67.90.142 --netmask=255.255.252.0
--nameserver=10.65.32.5 --hostname=onpcrfapp01.<thing>.org --device=eth1
--onboot=on
network --bootproto=static --ip=10.67.32.88 --netmask=255.255.255.240
--gateway=10.67.32.81 --nameserver=10.65.32.5
--hostname=onpcrfapp01.<thing>.org --device=eth0 --onboot=on
(mangle):
#if $gateway != "" and $iname != 'eth1':
#set $network_str = $network_str + " --gateway=" + $gateway
#end if
I'm really at a loss here.
Any suggestions, or should this go to the kickstart list?
10 years, 7 months
RFC: Migrating to GRUB2 for UEFI/PXE?
by Anthony Alba
Hello James and Cobbler folks.
This is an RFC for migrating to GRUB2 for UEFI/PXE netboot.
1. Background: Cobbler currently provides grub-{x86_64,x86).efi for netboot
for UEFI systems. This is based on the Red Hat-type-distros hacked
grub-legacy with UEFI net support.
By and large this works, but on my testing with Dell UEFI firmware it is a
bit hit-or-miss: it works with VNC installs but the console isn't always
initialized correctly. GRUB legacy isn't actively developed any more, so it
doesn't seem worthwhile to track down the root cause.
2. Proposal: to switch to GRUB2 for UEFI/PXE installs. GRUB2 is actively
developed, both upstream, and Fedora >= 19, and it has a config file search
algorithm similar to grub-legacy.
Fedora 19:
/grub2-efi/grub.cfg-01-30-85-a9-46-ae-9e
/grub2-efi/grub.cfg-C0A80196
/grub2-efi/grub.cfg-C0A8019
/grub2-efi/grub.cfg-C0A801
/grub2-efi/grub.cfg-C0A80
/grub2-efi/grub.cfg-C0A8
/grub2-efi/grub.cfg-C0A
/grub2-efi/grub.cfg-C0
/grub2-efi/grub.cfg-C
BTW upstream GRUB2 does not do this, and will require kssendmac
functionality to work correctly
I have found GRUB2 to work reliably with the same variety of Dell UEFI
bioses.
3. If people think this it worthwhile, I could prepare a patch,
side-by-side with grub-legacy. It will not affect existing setups and
require manual intervention to activate to achieve further testing.
BTW, this, of course, doesn't affect the pos-install OS boot loader. Even
if Cobbler is using GRUB2, at the end of a RHEL6, CentOS6 install the OS
will still be using grub(-legacy).
-A.Alba
10 years, 7 months
How do I specify virt-bridge with XMLRPC?
by Dylan Martin
Hi all.
I like to populate cobbler from a custom script and I've discovered that
it's not setting virt-bridge. I've done some poking around and tried a few
things but no luck.
Most network interface changes take the form of:
server.modify_system(system_id,'modify_interface', {
"macaddress-eth0" : "01:02:03:04:05:06",
"ipaddress-eth0" : "192.168.0.1",
"dnsname-eth0" : "hostname.example.com",
}, token)
I've tried 'virt-bridge' and 'virt-bridge-eth0' If I try
'virt-bridge-eth0' it complains and dies. If I try 'virt-bridge' it
doesn't complain, but it doesn't work either.
The VMs I'm trying to build with cobbler all have two interfaces which
attach to two different bridges. If it is just 'virt-bridge' and I'm doing
something else wrong, how do I tell it which interface to modify?
Thanks!
-Dylan
10 years, 7 months
