tore-'s questions

I'm currently writing a puppet module to automate the process of joining RHEL servers to an AD domain, with support for Kerberos.

Currently, I have problems with automatically obtain and cache Kerberos ticket-granting ticket via kinit. If this were to be done manually, I would do this:

kinit [email protected]

This prompts for the AD user password, hence there is a problem with automate this.

How can I automate this? I've found some posts mentioning using kadmin to create a database with the AD users password in it, but I've had no luck.

Trying to install collectd, but for some reason it claims python isn't installed:

yum install collectd
Loaded plugins: fastestmirror
Loading mirror speeds from cached hostfile
Setting up Install Process
Resolving Dependencies
--> Running transaction check
---> Package collectd.i386 0:4.10.0-4.el5 set to be updated
--> Processing Dependency: libxml2.so.2 for package: collectd
--> Processing Dependency: libcurl.so.3 for package: collectd
--> Processing Dependency: libpython2.4.so.1.0 for package: collectd
--> Processing Dependency: libidn.so.11 for package: collectd
---> Package collectd.x86_64 0:4.10.0-4.el5 set to be updated
--> Running transaction check
---> Package collectd.i386 0:4.10.0-4.el5 set to be updated
--> Processing Dependency: libpython2.4.so.1.0 for package: collectd
---> Package curl.i386 0:7.15.5-9.el5 set to be updated
---> Package libidn.i386 0:0.6.5-1.1 set to be updated
---> Package libxml2.i386 0:2.6.26-2.1.2.8 set to be updated
--> Finished Dependency Resolution
collectd-4.10.0-4.el5.i386 from epel has depsolving problems
  --> Missing Dependency: libpython2.4.so.1.0 is needed by package collectd-4.10.0-4.el5.i386 (epel)
Error: Missing Dependency: libpython2.4.so.1.0 is needed by package collectd-4.10.0-4.el5.i386 (epel)
 You could try using --skip-broken to work around the problem
 You could try running: package-cleanup --problems
                        package-cleanup --dupes
                        rpm -Va --nofiles --nodigest
[root@vitlinpup001 templates]# find / -iname "libpython2.4.so.1.0" -type f
/usr/lib64/libpython2.4.so.1.0
[root@vitlinpup001 templates]# file /usr/lib64/libpython2.4.so.1.0
/usr/lib64/libpython2.4.so.1.0: ELF 64-bit LSB shared object, AMD x86-64, version 1 (SYSV), stripped
[root@vitlinpup001 templates]# ls -lah /usr/lib64/libpython2.4.so.1.0
-r-xr-xr-x 1 root root 1.2M Sep  3  2009 /usr/lib64/libpython2.4.so.1.0
[root@vitlinpup001 templates]# yum info python | grep installed
Repo       : installed

I've tried to run package-cleanup --problems, package-cleanup --dupes and rpm -Va --nofiles --nodigest.

This is a first timer for me regarding missing dependency, so I'm a little lost on how to proceed.

I'm currently setting up RHN Satellite, and all works well. I'm in the process of creating custom channels, since we have certain software which should be available for all nodes of satellite, e.g. puppet, facter, subversion, php (newer version than present in base).

I've tried to find documentation on best practices on this. How should they be set up, how to handle different arch, how to handle noarch packages. How to sync updates to dependencies when updating a custom package in a custom channel (e.g. php is updated, how to fetch all updated dependencies).

The channel management documentation from RHEL (http://www.redhat.com/docs/en-US/Red_Hat_Network_Satellite/5.3/Channel_Management_Guide/html/Channel_Management_Guide-Custom_Channel_and_Package_Management.html) doesn't provide me with enough information on how to solve any of theese issues.

All tips, tricks and information regarding this would be great!

I've just setup a new production environment with puppet. Using 0.25.4 in client/server. Ruby is at 1.8.5, CentOS 5.4.

I've made a simple manifest for configuring yum-updatesd, but the puppetmaster doesn't seem to notice changes done to site.pp:

err: Could not parse for environment production: Could not match 'node' at /etc/puppet/manifests/site.pp:1
err: Could not retrieve catalog from remote server: Error 400 on SERVER: Could not parse for environment production: Could not match 'node' at /etc/puppet/manifests/site.pp:1

Notice, it says line 1. But line 1 contains an import statement:

# cat -n /etc/puppet/manifests/site.pp
1 import "update-notification"
2
3 node default {
4 include update-notification
5 update-notification::configure()
6 }

I've tried to reboot the server, delete and recreate site.pp, start and stop puppetmaster and puppet, with no luck. What am I missing?

I'm currently trying to figure out a smart way to set IP, hostname, gateway and DNS settings during a kickstart (with user input during the kickstart). Doing this with firstboot after the install is not acceptable, so this must be done during %post.

I've tried to run the system-config-network tool during post in tty6:

#!/bin/sh  
chvt 6  
exec < /dev/tty6 > /dev/tty6  
/mnt/sysimage/usr/bin/system-config-network-tui  

This doesn't work since for some reason I'm unable to catch user input.

I would prefer to use built in tools to change this during post, rather than write my own bash script to do this, since using the provided tools is less likely to break anything.

Does someone done anything like this or similar, and made it work?

Thanks

I'm currently writing a module for puppet which installs VMWare tools to virtual nodes.

I want to do this via yum and and yum-repo. VMWare have their own repo (http://packages.vmware.com/tools/esx/3.5latest/rhel5/x86_64/index.html) which I thought I could use, rather than creating my own. But then I noticed that their repo files is alot different than the rpm file used when installing VMWare Tools on the node, via the "Install/upgrade VMWare Tools" in vSphere.

Does anyone know what the real difference is? Does anyone have any preferences?

Currently testing cobbler, but have problem with the kickstart script when the partition information is loaded.

Here is my ks: http://pastebin.ca/1824343

I can't figure out what is the problem with the partsection at all. Without it, it works. I've even tried autopart. If the entry is removed, it works, but of course I have to provide the installer with partition information.

Under the kickstart an python exception is raised. I get a Errno 2 No such file or directory. My Apache logs states:

File does not exist: /var/www/cobbler/links/CentOS-5.3-x86_64/images/updates.img

File does not exist: /var/www/cobbler/links/CentOS-5.3-x86_64/disc1

File does not exist: /var/www/cobbler/links/CentOS-5.3-x86_64/images/product.img

But without the part information, no error occours.

What am I not seeing?

Cobbler 2.0.3, imported the CentOS 5.3 x86_64 DVD, PXE booting from a Xen guest.

Im currently building a cobbler and puppet setup. One thing I do is disabling unwanted services with puppet, but then I thought: Should I remove them with cobbler, rather than disabling them after they are intalled?

For example, my base class in puppet informs the client to disable smartd, cpuspeed, microcode_ctl, and readahead_early if its a virtual machine.

Why not remove then? If I do need some of these packages on a later stage, I could just install. Of course, if the machine is converted to a physical hardware, I would have to remeber to install these packages again.

I think that the best solution to my question is to create a puppet class which disables service A if its virtual, and enables it if its not. But, maybe someone else have some other insight?