FWIW,
(1) at my current shop, there's an immense hatred of everything JVM. That's
going to be a hard transition. Not to mention Puppet is the only place we
run Ruby, so it's nice and easy to let puppet do whatever it wants with
Ruby. Not so much for installing JVMs that may break production
Agents never control other agents. Aside from supporting technology
(PuppetDB, an ENC if you have one, a database to back PuppetDB), yeah, a
master and N agents is the gist of it.
There are some docs online (see specifically the Tuning and Scaling
section of the Puppet Documentation Index,
IMO...
The right way to do this is to use a parameterized class in your module
that generates the config files.
Beyond that, it's up to you how you set that parameter to the correct value
for a node - you could use Hiera, you could use an ENC, or if you still
have a small setup and are using
Hmm... I didn't even know this existed. Ironically, given your question, it
sounds like something I'd want to use. But if it's going away, I guess I'll
just totally forget that I heard it...
Use case: We still have a bunch of legacy systems that aren't puppetized
and probably never will be (lots
, right?
On Sat, Sep 27, 2014 at 9:07 AM, Jason Antman ja...@jasonantman.com
wrote:
Agents never control other agents. Aside from supporting technology
(PuppetDB, an ENC if you have one, a database to back PuppetDB), yeah, a
master and N agents is the gist of it.
There are some docs online
What facter versions are you running?
On Fri, Sep 26, 2014 at 6:02 AM, kaustubh chaudhari kaustu...@gmail.com
wrote:
Hi,
I am also facing same issue. unable to find where to look for, puppet
agent runes file facter runes fine if run manually.
But schedule run still not working.
Any
Not built-in to puppet, but this should be trivial to do with nmap or a
similar tool and a list of the nodes that *are* running puppet.
On Fri, Sep 26, 2014 at 10:35 AM, kaustubh chaudhari kaustu...@gmail.com
wrote:
Hi,
Just had a though if this is possible. I have seen this feature in HPSA.
IMO there's no sane way to do this automatically. What would it do, put
every file on the system into a module? The code certainly wouldn't be
properly modularized and parameterized, and would be a maintenance
nightmare.
On Thu, Sep 25, 2014 at 8:42 AM, Paul Tötterman paul.totter...@gmail.com
Yeah. I found an error in the puppetlabs-mcollective module (
https://tickets.puppetlabs.com/browse/MODULES-1192) where it doesn't work
with trusted_node_data = true, which per the docs is a 3.6.2 recommended
safe setting. When I was testing my fix, I made the mistake of running
specs with future
The few times I've had to do things like this (like pulling information
from a physical inventory database to find the rack a host is in, for
monitoring) I've just cached the fact value on disk, usually using some
sort of mtime-based TTL. That being said, I've only done this for facts
that are
FWIW,
1) for logging, we use ELK. We tried Splunk... the quote they gave us was
somewhere around our entire annual software budget. Also, it's closed
source.
2) We don't do anything with logs, aside from the few times we need to
manually confirm something. We rely on reports (PuppetBoard, and
Agreed with Atom...
I generally think that this method is backwards. The system shouldn't tell
Puppet what it wants to be; Puppet (possibly fed by some external data
source(s)) should tell the system what to be. Sure, in some very large
environments with a small number of possible configurations,
There's at least one open issue for this,
https://tickets.puppetlabs.com/browse/PUP-1376
It doesn't appear to have much interest, so I doubt it will get much
attention.
-Jason
On Thu, Aug 14, 2014 at 6:25 AM, Maciek Jackowski niksfire...@gmail.com
wrote:
is it possible to run agent with
Two things that I've come across that help this situation:
- for new modules, rspec tests
- for existing modules, https://github.com/ripienaar/puppet-catalog-diff -
I generate a catalog with the new and old branch, then diff them
On Fri, Aug 22, 2014 at 9:34 AM, Martin Langhoff
The module doesn't install git from source, because nobody should install
packages from source on a RedHat derivative OS. Have you read the
documentation for the git module? Or the package type, since you're
specifying the source parameter?
On Tue, Aug 12, 2014 at 7:17 PM, Vikas Kumar
Seb,
You really shouldn't be running specs for dependent modules.
(1) If I were you, I'd really use the spec helper, and garethr's module
skeleton - https://github.com/garethr/puppet-module-skeleton
(2) looking at your code, you should just be able to
t.pattern =
On Thu, Jul 31, 2014 at 9:21 AM, jcbollinger john.bollin...@stjude.org
wrote:
That being said... don't run via cron.
... if all you want it for is provisioning. For Puppet's core use --
ongoing configuration management -- there are a lot of advantages to
scheduling agent runs via cron
Yeah. You don't need to give them access to the master, you need something
(I use a git hook on the git server) that deploys new pushes to the master.
On Tue, Aug 5, 2014 at 12:15 PM, Christopher Wood
christopher_w...@pobox.com wrote:
Off the top of my head, have them commit/push to the same
So... there *is* a bug open that relates to this,
https://tickets.puppetlabs.com/browse/PUP-2144 (and by extension
https://tickets.puppetlabs.com/browse/PUP-1362 which appears to be being
worked on, and would fix this).
I don't know if that should work, and I honestly don't know how to set a
(1) a compile test is good, really good. make sure cheap tests happen first
(and fail fast) before expensive ones.
(2) I have a strong bias for having the automated jobs just run bundle
exec rake test or something like that, so that you can use the same job
for multiple modules
On Thu, Aug 7,
Seb,
I actually *just* came by a very similar issue myself. The examples at
http://murphyslaw.github.io/blog/2012/04/06/run-specs-excluding-integration-specs/
should help you.
-Jason
On Wed, Aug 13, 2014 at 8:42 AM, Jason Antman ja...@jasonantman.com wrote:
Seb,
You really shouldn't
I've seen those environments. I've worked in them. A few host types in my
current environment are like that. IT IS A BUG. The only valid reason for
this is either a bug in your manifests/modules, or that things aren't
ordered properly.
That being said... don't run via cron. If you're using any
Puppet is idempotent. It doesn't work that way. Puppet declares intended
state. So you tell puppet service should be running and files should be
there and Puppet makes it so. If you have a thing that should be
implemented/executed in a specific way with specific ordering, that's what
custom
Before you get any further, you do understand that the inifile has to be on
the *master*, right? Just checking, because I've seen a lot of people
trying to write functions, and only later realizing that the data they want
is on the node, not the master.
When I've come up with issues like this
Just in case anyone comes by this in the future, maybe I'm missing
something, but why not keep this really simple (did you read the type
reference docs for host?):
@@host { ${::hostname}_exported :
name= $::hostname,
ensure = present,
ip = $secondary_ip,
}
host {
Agreed with everything Schmitt said. This isn't what mcollective is for,
that's not how it works, there are tools made especially for this, etc. I'm
not saying it isn't possible if you jump through enough hoops. I'm just
saying it's an incredibly bad idea.
On Mon, Jul 14, 2014 at 2:41 PM, Andrii
For future reference...
(1) if at all possible (I assume you did this) hacking up upstream modules
is bad
(2) I'd use pip2pi (https://github.com/wolever/pip2pi) to create a mirror
of the modules you want, and then have puppet manage /root/.pip/pip.conf
(assuming puppet is running as root) and set:
I've used two different ways of doing this, depending on whether or not you
have this stuff installed/running:
(1) if you have PuppetDB, just use that. I probably have a (Python) check
plugin for this somewhere if you need it.
(2) use `mco puppet status` and parse the output of that
These also
I'm also experiencing this issue. I've opened
https://tickets.puppetlabs.com/browse/MODULES-1190 for it
--
You received this message because you are subscribed to the Google Groups
Puppet Users group.
To unsubscribe from this group and stop receiving emails from it, send an email
to
You should use the 'require' metaparameter, or if you absolutely need to,
explicit ordering. docs.puppetlabs.com will give you further information.
You should also read the Style Guide, and not throw all of that in one
manifest.
On Wed, Jun 25, 2014 at 8:04 AM, Malintha Adikari
FWIW, two thoughts on this:
1) Tunnel over SSH or VPN.
2) Is it feasible to *not* use a puppetmaster for this, and rather have
something on the laptops (cronjob? small daemon?) that pulls down your
config repository and runs masterless puppet locally?
-Jason
On Wed, Jun 18, 2014 at 10:40 AM,
Joseph,
See https://tickets.puppetlabs.com/browse/PUP-682
I'm going to try and get the pull request rebased, but at best this will
be in puppet4.
-Jason
On 06/12/2014 02:44 PM, Joseph Swick wrote:
Hi list,
I'm working on a little addition to an internal module we use to ensure
our puppet
That's not how it works.
In the output below, you can clearly see that Puppet is executing
`*/bin/rpm -i
puppet:///development/java_rpm/files/jdk-7u25-linux-x64.rpm*`. Why would
that work? puppet:/// means nothing to RPM, and puppet:/// is not a
valid 'source' for the Package type.
You
Yes, I have an idea.
1) Read the documentation:
http://docs.puppetlabs.com/guides/external_nodes.html
2) If that doesn't help, post your ENC code, with specific examples of
output and problems that you're having.
On 06/17/2014 07:49 AM, shlo.af...@gmail.com wrote:
Hi,
I understood ENC
I can think of 2 ways of doing this, none of which are terribly nice:
1) use a separate Environment with the new version of the module, and start
running nodes against that. Once you get them all moved over to the new
environment, roll out the new module version to the production environment,
and
?
Thanks,
Jason Antman
On Fri, Mar 21, 2014 at 2:16 PM, Heidi Pio heidi@puppetlabs.com wrote:
Hi Everyone,
Quick introduction: I'm the Engineering Project Manager of the Puppet
Labs Forge.
And a quick update to let you know that the Puppet Labs Forge and Module
issue migrations from
Matt,
I don't remember the technical details on why it exists, but it does.
No, there aren't any plans to relax it that I'm aware of...
Yeah, we use git branches as environments, and we essentially do
's/[^A-Za-z0-9_]/_/g' on the branch names to get environment names.
-Jason
On 02/19/2014
Agreed with what Matthias said. Puppet's naginator stuff doesn't play
very well with non-Puppet nagios configurations. Essentially, you have
to pick one of:
- remove the default template from that file, and create it with Puppet
elsewhere
- put all of the contents of that file in Puppet
-
There are some issues with Puppet restarting (or stopping) itself.
Beyond that, yes, I use puppet modules to manage the configuration of
all of my masters and agents.
There's a puppetlabs-puppet module, but it's horribly out of date. At
the moment, I'm using a fork
I seem to be pointing someone to this every week. You may want to refer
to John Bollinger's 2014-01-03 reply to the wondering if I want to[sic]
much now thread, which gives some excellent descriptions on 'how not to
Puppet':
https://groups.google.com/d/msg/puppet-users/IGqjPpVCrKA/VcUKiV3xfPkJ
by using the regular before and require. This should work but
now I have to consider a bunch of test scenarios.
I'll take a look at that rabbit class to see what they did
Thanks,
Dan
On Wed, Feb 5, 2014 at 12:32 PM, Jason Antman ja...@jasonantman.com
mailto:ja...@jasonantman.com wrote
Unnamed poster,
Succinctly: You're trying to write a bash script in Puppet. That's not
how it works, that's not what it's designed for, that's not how it
should be used.
However, luckily, there's a really simple and easy way to do what you're
trying to in Puppet:
Class['myapp::phase1'] -
On 02/05/2014 08:57 AM, dc12...@gmail.com wrote:
Thanks, I will take a look at that post.
I haven't been able to find the proper way to do some of these things.
Initially I had this running as a single class, working only with
requires, before, and notify.
This is where I ran into
Jon,
I've been using ENCs for quite a while, and have never used them with
site manifests (or any other manifests outside of modules). At the
moment, when I do things like that, I either do them inside a class
(i.e. put what you have in a wrapper class), or use require on all
resource
Disregard some of what I said. Sorry for the confusion. Stages can only
be used on classes not individual resources, so that won't work as
cleanly with forge modules (or anything that isn't specifically setup to
use stages).
On 02/04/2014 07:43 AM, Jason Antman wrote:
Jon,
I've been using
Steve,
I'll leave it up to others to answer your question more directly, as I'm
not sure I really can - it's been a while since I worked in an ITIL
shop. What I will say, though, since inevitably *someone* will, is that
ITIL contains some good concepts and some bad ones. Overall, I'd never
Your exec resource has a double-quoted string.
1) You're not doing variable interpolation, so that should be a
single-quoted string (per the accepted style).
2) Puppet *happens* to allow \#. The docs on Puppet string types and
escaping
references/resources, and I'll get back to you.
Thanks,
Jason Antman
PS - I hope the community guidelines are still correct, and job postings
are welcome here. I've seen very few, but we're hiring multiple
positions (scaling out development by about 20%, and the ops/automation
teams likewise), so we're
you have your own
existing talent do things, which sounds like your best bet:
http://www.slideshare.net/PuppetLabs/stophiringstartgrowing-130823160712phpapp01
On Sat, Feb 01, 2014 at 10:14:09AM -0500, Jason Antman wrote:
Is there a puppet jobs list or board anywhere? (not jobs at PL, jobs
dealing
, and having abysmal luck with wherever our internal and
external recruiters have posted.
Thanks for any tips/ideas,
Jason
--
Jason Antman | Systems Engineer | CMGdigital - Atlanta, GA, USA
jason.ant...@coxinc.com | p: 678-645-4155
--
You received this message because you are subscribed
That issue should be fixed soon, I'm told. The reports do get sent to
puppetboard, but the report processor throws an unhandled exception
because it references one of the 'metrics' hash fields, which is empty
in failed reports (this should be fixed in master to log an error
message instead).
For the reporting side, I'm using a Python project called PuppetBoard at
the moment - https://github.com/nedap/puppetboard - and it does
everything I liked about Dashboard, and also loads pages in a not just
reasonable but fast amount of time. It pulls directly from PuppetDB.
In terms of the ENC
I haven't used a dashboard version newer than... 2 years old or so...
but as of that version, and the version of Console that shipped with
Puppet Enterprise 2.5, no, this isn't possible. It's been a feature
request for years, and is one of the main factors that drove me to
develop my own ENC.
There's nothing existing that I know of that works in the GUI-based way
you seem to be talking about. Because, well, we *nix people usually
don't do that.
I've really only worked on RPM-based systems, so I'm not sure if this is
still applicable in the debian world...
There are 2 types of updates
it occurred is no longer known.
Someone else will probably chime in and correct me if my memory was
totally inaccurate.
-Jason
On 01/17/2014 07:19 AM, Andrea Cappelli wrote:
Il 17/01/2014 13:12, Jason Antman ha scritto:
On 01/16/2014 11:56 AM, Andrea Cappelli wrote:
Hi,
I'm searching through
Kaustubh,
The *easiest* method would be to use PuppetDB, but until a feature
request (https://tickets.puppetlabs.com/browse/PDB-16) is finished, it
only stores successful reports, not failed ones. Hopefully that ticket
will get closed soon, but it would probably be at least weeks until it
gets
James,
I vaguely remember seeing this 'node_terminus called twice' thing in the
past. The simple answer, though I know it's not what people want to
hear, is that the ENC should always return the right information. If you
want to modify the content of the catalog based on something that
happens
This is one of the best descriptions of how to do things in Puppet (or
maybe how not to) that I've seen, and touches on a *lot* of common
misunderstandings.
I think I might have to keep a short url to the archive of this handy...
Thanks, John.
On 01/03/2014 10:35 AM, jcbollinger wrote:
On
I *think* that Empty response is the issue here... your script appears
to be sending back YAML with just an (empty) classes hash and an (empty)
parameters hash. Try putting something in parameters (at least). Also
make sure that it's actually writing to STDOUT, not STDERR.
Assuming neither of
After digressing into this discussion on #puppet last night, I was
wondering what the community feelings are on ENCs vs Hiera...
I know that Dashboard/Console still exists, but have heard rumors (for
years) of it being either replaced by something else, or totally
rewritten. Then there's Foreman,
I guess I hit a bit of a nomenclature issue here. By ENC I didn't just
mean the node_terminus script, but the whole app - i.e. ENC as in the
way Dashboard is an ENC...
Though your comment does remind me that Hiera can now be more than a
directory full of YAML files... I guess I'll start doing
which need it, including Puppet.
I take it (hopefully) this is not too unusual and bizarre in the world
of Puppet.
On Thursday, December 5, 2013 4:16:10 AM UTC-8, Jason Antman wrote:
PuppetDB isn't an ENC. PuppetDB does, however, use Postgres
(unless you use the embedded database
Hello,
I've recently learned that my plans to use the puppet agent mcollective
plugin to trigger one-time runs against a different environment, with
the daemon running in the background, don't work because of how the
agent plugin works (SIGUSR1 to the daemon if running).
My original intent was
I also forgot the scariest option, which seems apt to break things:
- have mcollective stop the daemon
- mco puppet runonce
- have mcollective start the daemon back up
-jason
On 12/31/2013 07:42 AM, Jason Antman wrote:
Hello,
I've recently learned that my plans to use the puppet agent
* * * * /usr/bin/mco puppet runonce --noop --splaylimit 900
I don't know if it's the best solution for productive-environment,
opinions?
regards,
El martes, 31 de diciembre de 2013 13:50:11 UTC+1, Jason Antman escribió:
I also forgot the scariest option, which seems apt to break things
Thanks for the lengthy explanation. I suppose that, for one thing, I
didn't understand the algorithm that was used in puppetcommander. The
one thing that will complicate this for me is my horribly poor ruby
skills, coupled with the fact that the MCo python bindings appear to be
abandoned (though I
Since you're asking about a Ruby IDE, not a Puppet IDE, perhaps this
question would be better asked of a Ruby list, or even better, a search
engine...
On 12/24/2013 06:21 PM, Stuart Cracraft wrote:
Hi,
Is there a plain-text visual debugger ide for Linux for Ruby anyone
can mention?
Stuart
Peter,
I looked at puphpet a bit, and had it generate a sample config for me. The
puppet manifests that it outputs are VERY far from anything that could be
called best practices. The 'all one .pp file' paradigm leads me to believe
they're not even intended to be read or modified by a human,
Harshita,
That error message sounds like your certificate problem is with
PuppetDB, not the agent to the master.
the master is learn.localdomain.pem and the windows agent is the one
with the utterly useless hostname (igtggn*)?
Every node (master and agent alike) will always generate their own
Michael,
Yes, package resources (like all resources) need to be checked before
puppet can do anything with them. I'm not quite sure what you mean by
use yaml with a single command - puppet Package resources must be
specified with one single package name, there's no way around this that
I'm
Stuart,
You have *very* specific requirements here. In these situations, there
are generally three options:
1) find someone else who's done it and use theirs
2) learn how to do it yourself
3) pay someone else to do it
4) don't do it
You posted your question (which I assume you researched before
I've gone through and migrated most of the issues I was watching (and
still care about) that weren't already.
I ran into 4 issues that show up with a message of:
This issue is currently not available for export. If you are
experiencing the issue described below, please file a new ticket in
It's great that you're using Puppet and MCollective and trying to get
them setup everywhere, and there are some really easy ways to do that
when provisioning a new VM - either through templates or the build
process. Personally, I have Puppet baked in to our base installs (we
actually don't use
Linux apps which need it, including Puppet.
I take it (hopefully) this is not too unusual and bizarre in the world
of Puppet.
On Thursday, December 5, 2013 4:16:10 AM UTC-8, Jason Antman wrote:
PuppetDB isn't an ENC. PuppetDB does, however, use Postgres
(unless you use the embedded
PuppetDB isn't an ENC. PuppetDB does, however, use Postgres (unless you
use the embedded database, which you shouldn't). Puppet Dashboard is an
ENC, but ironically, uses MySQL not Postgres.
Stuart,
Starting *another* ENC thread a day later isn't likely to get you many
more responses than the
don't work this
way) or only when manually requested...
So I suspect things aren't so simple for you. I just don't see in what
manner.
Thanks in advance for any clarification,
Felix
On 12/03/2013 02:39 AM, Jason Antman wrote:
Hello,
I have 3 puppet stacks (master, puppetdb, enc) - dev, test/qa
Stuart,
Yes, many of us use ENCs, though it seems to be mostly limited to larger
shops; these days many people are moving in the direction of hiera
instead. I've been using various ENCs for the past ~4 years.
There are relatively few good choices for already-written ENCs. (well,
if you just want
, and when I want
to update the (exported resource) monitoring configuration for one of
the dev or test nodes, have to do a one-time run on each node in
question against the prod puppet master.
Any other thoughts or theories?
Thanks,
Jason Antman
--
You received this message because you
David,
What distro/OS are you running on? Have you tried looking
in$APP_DIR/log/app.log ?
That init script is not compliant with the Fedora/RedHat spec, so I'm
not sure if that would have anything to do with it (or, if you're on a
modern version of a Fedora/RHEL derivative, systemd/upstart
As far as I know, /usr/bin/puppet_node_classifier is supposed to be
the path to your ENC script.
What ENC are you trying to use? Either your ENC documentation should
tell you how to configure this (Puppet Dashboard/Console, Foreman, etc.)
or if you're writing your own ENC, this should be the
I'll wager that Michael is correct here (based on what he says, not just
his knowledge of packaging).
Can you provide the actual, full names of the package in question, or
the rpm -qi output?
My guess, based on having the same problem, is that if
ensure = '5.35.0-3_el5'
is failing, then you'll
I've got to agree with JuanBrein, I've been using puppet since 0.24.5
and I don't remember ever seeing a problem like this. How are you
deploying your modules?
Even if you have a time consuming deploy process (i.e. not just a git
fetch and pull), and inconsistent state on disk is a valid concern,
on a single canonical module for a given use case or package.
-jason antman
On 11/19/2013 02:48 PM, Matt Simmons wrote:
Hi John,
I'm new around here, but I'm also in the same situation as Tom, who
started this thread.
I was wondering if you could expound a little bit on the better
The right way would probably be to query PuppetDB if you're using it.
If you're not using it, the `puppet facts` face (at least in terms of
querying by node) seems to just be a CLI for the Inventory Service REST
API, which is documented at
I haven't read the new edition, but the last was good.
As much as I hate to say this (for all of the people who write Puppet
books) I've found that things are changing fast enough that I usually
recommend one or two books for a beginner to get the ropes, and then
after that, rely on the
Tom,
I'm not sure I totally understand your question; I've also never used an
ENC (like dashboard) with default.pp, only with modules.
If you set a parameter called env on a group in Dashboard, and then
add node foo.example.com to that group, the parameter will available
in puppet as a
' to that same group, the variable will be '::env' or
'::bar' ?
Thx
Tom
Le mercredi 13 novembre 2013 13:29:44 UTC+1, Jason Antman a écrit :
Tom,
I'm not sure I totally understand your question; I've also never
used an
ENC (like dashboard) with default.pp, only
You definitely *can* move PuppetDB to another host, but I think you'd
either be mostly on your own doing it, or need to reach out to support.
AFAIK it's not supported by the installer, and I don't remember seeing
documentation on it.
The hardest part would probably be moving PuppetDB's database,
, Jason Antman wrote:
[snip]
package installation - but the module (or at least our version of it)
doesn't handle requirements files, and uses a define to pip install
packages, so a given package can only be installed in one venv on a
node.
I have a hacked-up python virtualenv module, since I
We use largely the same solution as Jo but with an ENC.
Packages that we really don't care about are usually just ensure =
present, so every machine built from a given release should have the
same version (upgrades because of dependencies aside), or ensure =
latest in the rare case that we've
Tom,
I've actually been working with similar issues lately (and am in the
process of working on a virtualenv module).
I have a python module that includes classes for the common dependencies
(i.e. require python::pyyaml) and have been pretty happy with that
pattern so far, but if you want, I
interested in finding
something suitable that could become the official module.
Thanks for any advice, input or suggestions,
Jason Antman
jantman
(ccing djmitche as I can't find a mozilla releng email list)
--
You received this message because you are subscribed to the Google Groups
Puppet Users
Derek,
In most circumstances, yes, you should be running puppet commands as
root (via sudo). Running via sudo seems to be the standard, and the best
practice, in Linux environments. You could use some other methods if you
have a... unusual... environment, but running puppet commands as a
normal
Steven,
Can you be a little more specific about what you're trying to do?
Normally, you'd set that department variable in the ENC itself.
Puppet calls the ENC script (node_terminus) with the certificate name
(or is it the FQDN? I never remember, since they're both the same for
me) of the node
supports
- it provides one place to go to extend the module for new OSes
-Jason Antman
On 10/24/2013 06:13 AM, DEGREMONT Aurelien wrote:
Hello all,
As said in the subject, I did not find in puppet doc the official
recommendation in how multi-os should be handled in module manifests.
Let's
the same values in
every place,
or,
b) I need to (yuck) build pre-YAML-serialization variable interpolation
*into my ENC*...
Thanks for any advice,
Jason Antman
--
You received this message because you are subscribed to the Google Groups
Puppet Users group.
To unsubscribe from this group
missed, or will this end up being some
script that I cobble together myself?
Thanks for any advice/input
Jason Antman
--
You received this message because you are subscribed to the Google Groups Puppet
Users group.
To unsubscribe from this group and stop receiving emails from it, send an email
to believe that this is just caused by compiling a single catalog
on the master... Any advice or insight would be greatly appreciated.
Thanks,
Jason Antman
Trace from master:
/usr/lib/ruby/1.8/timeout.rb:64:in `rbuf_fill'
/usr/lib/ruby/1.8/net/protocol.rb:134:in `rbuf_fill'
/usr/lib/ruby/1.8/net
Commented out storeconfigs/thin_storeconfigs as we were only using them to
get facts in a DB and for Dashboard. Also commented out the Inventory
Service stuff. Looks like that did it. Now I just need to figure out the
real fix...
On Friday, September 14, 2012 10:40:16 AM UTC-4, Jason Antman
On 08/29/2012 02:09 PM, Jakov Sosic wrote:
On 08/29/2012 07:34 PM, Jason Antman wrote:
but given the way things work
around here, and that puppet is really a lower-level project (i.e.
experimentation, failure, and bugs aren't tolerated well because we only
have management support for the end
1 - 100 of 116 matches
Mail list logo
