similar to: No subject

Hi, I''m still writing my djbdns module, I came to the following design issue with master and slave tinydns. Normally there is no such master and slave distinction in djbdns: all tinydns instance are equal. Usually one instance on one host is considered the master on, which the RR are changed, and when you need to propagate them, you rsync the datas to the other hosts, where they will be

This appears to be working for me, but suggestions on improvements are welcome: master/manifests/definitions/stowedpackage.pp: > define stowedpackage ( $basepackage, $version, > $rsyncserver=''gold.cae.tntech.edu'', > $rsyncmodule=''metastow'', > $stowdestdir=''/usr/local/stow'' ) { > > file { stow_initiator:

I won''t claim it''s all done to the best style, but it does appear to work. Can someone take a look at http://blogs.cae.tntech.edu/mwr/2007/08/02/the-new-file-server-puppet-and-modules/ and give suggestions? It''s a bit more extensive than the example on the wiki at ModuleOrganization, and might make a decent jumping-off point for others. -- Mike Renfro / R&D

Lots of my puppet hosts have DHCP-assigned addresses. I''ve found, at least in Debian 4.0, that puppetd will often start up before the network interfaces are all up and running, and therefore it can''t contact the puppetmaster. When that happens, puppet exits immediately. I''m hacking around the problem with the following content in /etc/default/puppet, but

I''ve been working up my first puppet module, one that will make Amanda configurations consistently. In general, I''d like the following to happen: - Each Amanda server should make sure all its holding disks are configured properly; this should only happen once per server, per puppetd run. Definitely not on a per-configuration basis. - Each Amanda server should keep its

My first public-facing deployment of Puppet is likely to be on some parallel computing nodes we have here. One of our fluid dynamics packages expects to be able to rsh from one master node to other slave nodes in the same batch queue. I think I''ve made some good class definitions up to this last point, where code duplication rears its head. Example of the problem: class ch405-host

I''m trying to get a new file server managed by puppet from day 1, at least as much as possible. At the moment, though, there''s two issues I''m running into: 1. fstab should have entries for my comically-large RAID, but doesn''t. 2. each puppet run appears to remount the RAID, even when no rules in the manifest change. I suspect the issue may be in parsing

Is it possible to trigger a service restart without always forcing that service to be running? We shutdown mysql for about 30min every day to do backups, and I don''t want puppet to start mysql during this window. So I''ve removed ensure => "running" from the mysql service. But now notify => Service[mysql] doesn''t restart mysql. Is it possible to tell

I''ve converted all of the inline docs over to restructured text, and I''ve got them all up on trac: http://reductivelabs.com/trac/puppet/wiki/TypeReference http://reductivelabs.com/trac/puppet/wiki/FunctionReference http://reductivelabs.com/trac/puppet/wiki/ReportReference http://reductivelabs.com/trac/puppet/wiki/ConfigurationReference Please let me know if you find any

hi,all I want the puppetd run in daemon only run by the puppetrun trigger, not schedule run . so how can do this ? what''s the maximum of the runinterval ? I want change this option to do this. Please help me . -- Huang Mingyou

Not that I''m much of a Solaris guy, and definitely not a Solaris 10 guy, but since nobody else had posted anything yet: http://reductivelabs.com/trac/puppet/wiki/PuppetSolaris#Solaris10 The good news is that it appears to be a completely hands-off puppetd installation. Just sign the client key afterwards on the puppetmaster and you''re set. The bad news is that it''s

Hi, I''m absolutely new to centralized server management, sorry sorry in advance for any stupid remarks. Looking on the puppet documentation, I mainly find a lot of info about how to manage the config files, but nothing about centralized control. So I''m wandering if I''m looking a the right software package for my needs. I want to be able to do the following from a

Hello, I am examining whether puppet meets our needs. I can''t find the answer to two questions I am looking to answer, hopefully someone can help. 1. Can puppet be used to manage different classes of the same computers, e.g. do something on the apache test farm, if it works then do it on the apache live farm; and is this done cleanly? 2. How are package updates (e.g. yum update/yum

hi all , i want to disable runinterval option on puppetd . as i know , after every 1800 sec.( 30 mins) , puppet client pull configuration from puppet master . I want configuration to be pushed to puppet client when i wwould fire puppetrun on puppet master rather than after every 30 min , configuration pulled puppet client. runinterval :How often puppetd applies the client

I''m very new at puppet, but I''ve almost managed to get to the point where my puppet server sets up all the basics of a new linux box for me. The last piece of the puzzle is backups. The way I do backups is fairly simplistic. I just use mysqldump and tar to create local backup files and then scp them to a central host once a day. In order to set this up, I have to create a key

I tried to install the ''mongrel'' gem tonight via puppet in an attempt to migrate from webrick to mongrel as described on the puppet site[1]. I added the following to my manifest: package { "rubygems": ensure => installed; "mongrel": ensure => installed, provider => gem, require => Package["rubygems"]; }

Hi all, I''m a new puppet user and I''m running into some weirdness around creating users on solaris. (puppet version 0.23.2) The virt_all_users way of managing users seems pretty reasonable, but I don''t see how to deal with user specific groups with that. I have: class virt_all_users { @user { "seph": ensure => "present",

Hi, I have a template file which I would like to append to an existing file (in this case I wish to add a few lines to /etc/motd). Unless I''ve missed it, it does not appear that file has an append option. Is there an easy way copy to the end of a file Thanks Paul --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google

Hi, I am fairly new to using puppet, but have it setup on 25 machines. Just doing some basic configuration management. Recently I decided to try working with updating some services config files and used the ensure => running tags. My question is simple - if the source file does NOT change and I use a subscribe, why does the system keep restarting the daemon every time puppet runs? A simple

So I have two classes for openvz setup. The first is openvz::setup and it sets up the box for OpenVZ.. then another class (in same .pp file) called openvz::master sets up the config for the main system (as opposed to a VE). The error I see is that the openvz::master class is executed (and fails) becuase it doesn''t seem to load up the setup class? debug: Calling puppetmaster.getconfig