How does Puppet Push Model works

[megha sharma]

Hello,

How does puppet push model works for application deployment. It should be done in such a way that deployment happens only when required and the same should be initiated from puppet master.

Thanks,

Megha

[Dirk Heinrichs]

Am 17.08.2016 um 05:14 schrieb megha sharma:

How does puppet push model works for application deployment. It should be done in such a way that deployment happens only when required and the same should be initiated from puppet master.

It's not really push. Puppet agent runs on each node and connects to the master every 30 minutes (default), asking for its configuration. So you need to change the nodes configuration on the master, so that it gets it next time it connects.

To simulate a push, you'd need to setup mcollective on both master and nodes. You can then initiate an agent run remotely, from the master.

HTH...

    Dirk

--

Dirk Heinrichs, Senior Systems Engineer, Engineering Solutions
Recommind GmbH, Von-Liebig-Straße 1, 53359 Rheinbach
Tel: +49 2226 1596666 (Ansage) 1149
Email: d...@recommind.com
Skype: dirk.heinrichs.recommind
www.recommind.com

[Rich Burroughs]

Well there is Application Orchestrator if you are using PE, but I think it is still pretty green. It's meant to be the push solution for deploying apps but I'm under the impression that not a lot of people are using it yet.

Mcollective is one option, for sure, but there are others. Most orchestration tools will work, like Ansible. I've also used the Python-based tool called Fabric.

At my current job we do this with Jenkins. A Jenkins job kicks off a shell script on the target host that runs "puppet agent -t" at the proper time.

That's really the model. You need some tool that will kick off the agent runs on the right target hosts at the right time. I think eventually the "puppet job" command that showed up with App Orchestrator will be able to do this too.


Rich

--
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 puppet-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/puppet-users/4ac46dcb-9957-0d14-93f2-72775868e45a%40recommind.com.
For more options, visit https://groups.google.com/d/optout.

[Kong Eng]

you could do the following (i've have not tested):
- set puppet agent to use cache catalog
- use mco to force a puppet run (use non-cached catalog)

The above method should give you greater control when and who gets the new manifests.

[Martin Alfke]

Hi Megha,

> On 30 Sep 2016, at 03:05, Kong Eng <kong...@gmail.com> wrote:
>
> you could do the following (i've have not tested):
> - set puppet agent to use cache catalog
> - use mco to force a puppet run (use non-cached catalog)

this is the FOSS way.
Within PE you also have the possibility to use the orchestrator.
The orchestrator uses direct puppet (pxp-puppet agent).
https://docs.puppet.com/pe/latest/app_orchestration_overview.html

>
> The above method should give you greater control when and who gets the new manifests.
>
> On Tuesday, August 16, 2016 at 11:14:48 PM UTC-4, megha sharma wrote:
> Hello,
>
> How does puppet push model works for application deployment. It should be done in such a way that deployment happens only when required and the same should be initiated from puppet master.
>
> Thanks,
> Megha
>

> --
> 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 puppet-users...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/puppet-users/24df1915-7912-4df2-bd6c-8bd5cde236c2%40googlegroups.com.