You are viewing a plain text version of this content. The canonical link for it is here.
Posted to infrastructure-dev@apache.org by Nick Kew <ni...@apache.org> on 2016/07/08 12:19:51 UTC

Planet at Apache

Please correct me if I'm wrong, but since the demise of
planet apache, I gather there's no appetite in the infra
team to support it.

Would it be any use if I take on the job of resurrecting it?
That would imply access to a server, and the relevant data
(is/was that fed from id.apache.org)?

I'd also be interested in whether planet could be extended
to aggregate per-project blogs, where a feed is available
that filters on a [projectname] keyword.  That could then be
offered to projects interested in aggregating project blogs.
Does that sound reasonable?

-- 
Nick Kew

Re: Planet at Apache

Posted by Sam Ruby <ru...@intertwingly.net>.
On Fri, Jul 8, 2016 at 11:58 AM, Sam Ruby <ru...@intertwingly.net> wrote:
> On Fri, Jul 8, 2016 at 11:27 AM, Chris Lambertus <cm...@apache.org> wrote:
>>
>>> On Jul 8, 2016, at 5:19 AM, Nick Kew <ni...@apache.org> wrote:
>>>
>>> Would it be any use if I take on the job of resurrecting it?
>>> That would imply access to a server, and the relevant data
>>> (is/was that fed from id.apache.org)?
>>
>> We’re happy to help facilitate this. Please put in an Infra jira ticket for the creation of a new puppetized VM. As long as the system is built within our puppet and security framework, we’re more than happy to have volunteers step up for the effort in resurrecting and maintaining this service.
>
> Any objections to putting it on comdev-vm?
>
> The relevant data needed is a list of feeds to subscribe to.

More information:

previous "runbook":
https://svn.apache.org/repos/infra/infrastructure/trunk/docs/services/planet.txt

Lists of lists of feeds:

http://svn.apache.org/repos/asf/planet/

>>> I'd also be interested in whether planet could be extended
>>> to aggregate per-project blogs, where a feed is available
>>> that filters on a [projectname] keyword.  That could then be
>>> offered to projects interested in aggregating project blogs.
>>> Does that sound reasonable?
>>
>> I suppose you’re welcome to extend the codebase. I’m not familiar enough with it to know how feasible this is.
>
> Here is the codebase:
>
> https://github.com/rubys/venus
>
>>>
>>> --
>>> Nick Kew
>>
>> -Chris
>
> - Sam Ruby

- Sam Ruby

Re: Planet at Apache

Posted by Chris Lambertus <cm...@apache.org>.
> On Jul 8, 2016, at 8:58 AM, Sam Ruby <ru...@intertwingly.net> wrote:
> 
> On Fri, Jul 8, 2016 at 11:27 AM, Chris Lambertus <cml@apache.org <ma...@apache.org>> wrote:
>> 
>>> On Jul 8, 2016, at 5:19 AM, Nick Kew <niq@apache.org <ma...@apache.org>> wrote:
>>> 
>>> Would it be any use if I take on the job of resurrecting it?
>>> That would imply access to a server, and the relevant data
>>> (is/was that fed from id.apache.org <http://id.apache.org/>)?
>> 
>> We’re happy to help facilitate this. Please put in an Infra jira ticket for the creation of a new puppetized VM. As long as the system is built within our puppet and security framework, we’re more than happy to have volunteers step up for the effort in resurrecting and maintaining this service.
> 
> Any objections to putting it on comdev-vm?


No objections, as long as comdev doesn’t mind.

Re: Planet at Apache

Posted by Sam Ruby <ru...@intertwingly.net>.
On Fri, Jul 8, 2016 at 11:27 AM, Chris Lambertus <cm...@apache.org> wrote:
>
>> On Jul 8, 2016, at 5:19 AM, Nick Kew <ni...@apache.org> wrote:
>>
>> Would it be any use if I take on the job of resurrecting it?
>> That would imply access to a server, and the relevant data
>> (is/was that fed from id.apache.org)?
>
> We’re happy to help facilitate this. Please put in an Infra jira ticket for the creation of a new puppetized VM. As long as the system is built within our puppet and security framework, we’re more than happy to have volunteers step up for the effort in resurrecting and maintaining this service.

Any objections to putting it on comdev-vm?

The relevant data needed is a list of feeds to subscribe to.

>> I'd also be interested in whether planet could be extended
>> to aggregate per-project blogs, where a feed is available
>> that filters on a [projectname] keyword.  That could then be
>> offered to projects interested in aggregating project blogs.
>> Does that sound reasonable?
>
> I suppose you’re welcome to extend the codebase. I’m not familiar enough with it to know how feasible this is.

Here is the codebase:

https://github.com/rubys/venus

>>
>> --
>> Nick Kew
>
> -Chris

- Sam Ruby

Re: Planet at Apache

Posted by Sam Ruby <ru...@intertwingly.net>.
On Tue, Jul 12, 2016 at 8:34 AM, Nick Kew <ni...@apache.org> wrote:
> On Fri, 8 Jul 2016 08:27:11 -0700
> Chris Lambertus <cm...@apache.org> wrote:
>
>> We’re happy to help facilitate this.
>
> In the light of Sam's suggestion that it should run on
> comdev-vm, I guess what I'd need is shell access to that
> rather than any new VM?

Possibly desirable, but not required.  The preferred way to set that
up would be through puppet.  I can help with that, though I may not
have cycles to start until Mondayish.  See
https://issues.apache.org/jira/browse/COMDEV-201 for an example.

> Is that also a matter for a servicedesk request?
> I've no idea what 'component' it should go under!

The general way in which changes are made to project VMs is that pull
requests are submitted against infrastructure-puppet and reviewed and
merged by infrastructure staff:

https://cwiki.apache.org/confluence/display/INFRA/Git+workflow+for+infrastructure-puppet+repo

If you are so inclined, I also am working on a tool that will help you
deploy a VM (such as comdev-vm) on a machine you host.  This enables
you to test out your changes:

https://github.com/rubys/infrastructure-playpen

I'm quite willing to set up planet for others to maintain, though even
better would be helping teach others how to set things up for
themselves.

> Thanks,
>
> --
> Nick Kew

- Sam Ruby

Re: Planet at Apache

Posted by Nick Kew <ni...@apache.org>.
On Fri, 8 Jul 2016 08:27:11 -0700
Chris Lambertus <cm...@apache.org> wrote:


> We’re happy to help facilitate this.

In the light of Sam's suggestion that it should run on
comdev-vm, I guess what I'd need is shell access to that
rather than any new VM?

Is that also a matter for a servicedesk request?
I've no idea what 'component' it should go under!

Thanks,

-- 
Nick Kew

Re: Planet at Apache

Posted by Chris Lambertus <cm...@apache.org>.
> On Jul 8, 2016, at 5:19 AM, Nick Kew <ni...@apache.org> wrote:
> 
> Would it be any use if I take on the job of resurrecting it?
> That would imply access to a server, and the relevant data
> (is/was that fed from id.apache.org)?


We’re happy to help facilitate this. Please put in an Infra jira ticket for the creation of a new puppetized VM. As long as the system is built within our puppet and security framework, we’re more than happy to have volunteers step up for the effort in resurrecting and maintaining this service.

> I'd also be interested in whether planet could be extended
> to aggregate per-project blogs, where a feed is available
> that filters on a [projectname] keyword.  That could then be
> offered to projects interested in aggregating project blogs.
> Does that sound reasonable?

I suppose you’re welcome to extend the codebase. I’m not familiar enough with it to know how feasible this is.

> 
> --
> Nick Kew



-Chris

Re: Planet at Apache

Posted by Sam Ruby <ru...@intertwingly.net>.
On Wed, Jul 13, 2016 at 10:10 AM, Nick Kew <ni...@apache.org> wrote:
> Sam:
>
>> The general way in which changes are made to project VMs is that pull
>> requests are submitted against infrastructure-puppet and reviewed and
>> merged by infrastructure staff:
>
> I see.
>
>> If you are so inclined, I also am working on a tool that will help you
>> deploy a VM (such as comdev-vm) on a machine you host.  This enables
>> you to test out your changes:
>
> OK, I had imagined "apt-get install planet-venus" created
> a local playpen for running it, and that a source install
> from planetplanet did the same for hacking it.  How much
> difference does the VM environment affect that?

Wrong question :-)

Puppet in a nutshell: you describe the state you want the machine in
question to be in, and puppet makes it happen.

For example, you want planet-venus installed as a package?  Here's
what you would type:

  package { 'planet-venus':
     ensure => installed
  }

Or suppose you wanted to check out the source instead:

  vcsrepo { '/home/applanet/git/venus':
    ensure   => latest,
    provider => git,
    source   => 'https://github.com/rubys/venus.git',
  }

Other things you will want to do (create an applanet user, create the
'/home/applanet/git' directory, add cronjobs, configuring a vhost with
apache httpd) are done similarly.  If there are ordering dependencies,
you can express them too using puppet.

Here's the quick reference for puppet:

https://docs.puppet.com/puppet/latest/reference/type.html

Things like vcsrepro are provided by third party modules.

>> I'm quite willing to set up planet for others to maintain, though even
>> better would be helping teach others how to set things up for
>> themselves.
>
> Thanks.  That sounds like an offer I can't possibly refuse!
> Two learning curves (Planet and ASFInfra) for the price of one.

It is more than two, but on the plus side you have people who can
answer your questions and point you in the right direction.

> I shall be sure to check out all your links within this week.

I'd recommend starting with
https://github.com/rubys/infrastructure-playpen.  Setting up a new
virtual machine for the first time will take about 20 minutes.  Once
done, you can "vagrant ssh" into that machine and look around.

Next, you are going to want to create an 'applanet' module in
https://github.com/apache/infrastructure-puppet/tree/deployment/modules.
Your puppet code will go into
infrastructure-puppet/modules/applanet/manifests/init.pp.

Finally, you will need to add the name of your module (applanet) to
infrastructure-puppet/data/nodes/comdev-vm.apache.org.yaml.  This goes
in the list of classes at the top of the file.  If you keep your
module self contained, moving planet to a different vm would be as
simple as adding that one line to the desired VM.  More likely,
however, is that your Apache vhost definition will go into the
comdev-vm.apache.org.yaml file, and those lines would also have to be
moved too.

Applying your changes will take about two minutes.  This is mostly a
function of the complexity of the ASF standard configuration.  I've
puppetized my development machines and server, and applying changes
takes about 20 seconds for my setup.

> --
> Nick Kew

- Sam Ruby

Re: Planet at Apache

Posted by Nick Kew <ni...@apache.org>.
Sam:

> The general way in which changes are made to project VMs is that pull
> requests are submitted against infrastructure-puppet and reviewed and
> merged by infrastructure staff:

I see.

> If you are so inclined, I also am working on a tool that will help you
> deploy a VM (such as comdev-vm) on a machine you host.  This enables
> you to test out your changes:

OK, I had imagined "apt-get install planet-venus" created
a local playpen for running it, and that a source install
from planetplanet did the same for hacking it.  How much
difference does the VM environment affect that?

> I'm quite willing to set up planet for others to maintain, though even
> better would be helping teach others how to set things up for
> themselves.

Thanks.  That sounds like an offer I can't possibly refuse!
Two learning curves (Planet and ASFInfra) for the price of one.

I shall be sure to check out all your links within this week.

-- 
Nick Kew