Platform Team Week of October 6th, 2014

Andy Parker

unread,

Oct 9, 2014, 3:24:30 PM10/9/14

to puppe...@googlegroups.com

** Next PR Triage Wednesday, October 15th @ 10:00 am Pacific. **

Priorities

Puppet 3.7.2
CFacter on the march
New puppet doc implementation
Code removal for puppet 4

Commentary

Puppet 3.7.2 is in full swing. The big unknown for me was the memory leak (PUP-3345) that had been reported, but we hadn't reproduced. Well, it's been reproduced. It happens when directory environments are in use and environment_timeout != unlimited. Whenever it refreshes the environment cache it seems to leak memory. This is pretty critical for us since it will be in the next PE version, work on which is taking a lot of our time at the moment.

CFacter is continuing. Solaris support is still in the works and so is the toolchain for building it on all the various platforms. I believe you can see the build scripts at https://github.com/puppetlabs/cfacter-build

Puppet-dev conversations of note:

* PUP-1073 - How to allow installing packages with the same name in different packaging systems in a single catalog?

This thread had been going on for a long time and had died off in March without a conclusion. The issue was brought back to my attention this past week and so I started the conversation up again with another proposal. There are some very valid concerns about what impact loosening the constraints will have on catalogs and flapping resources; however, it appears that the decision is to go ahead with the latest proposal. I still need to reply to the thread to conclude it and update tickets, etc, etc. I'm going to target a fix a 4.0.

* Behavior of apply + ENC

Felix presented some options about how command line arguments should behave wrt to environments + enc + command line overrides. There was a lot of agreement about a "command line args win" solution, but RI feels that "ENC wins" is the better solution. I'm waiting with baited breath for the conclusion :)

* Removing types/providers from core

Ah yes. This old thread. But it's going somewhere this time! Daenny and Kylo submitted a PR for removing nagios from core at the contributor summit and now Daenny is following up with working out a plant for removing the rest. I'd really like to see this drive to a conclusion that we take action on this time. At the moment a lot of things hinge on getting packaging setup to allow pulling in modules so that they don't live in the core repo, but are available in a standard puppet installation.

Puppet strings 0.1.0 has been released! Hailee put a lot of work into this to get it ready and Charlie did a lot of great work to show that it can be done. The next steps are for people to try it out and send us feedback/fixes.

Whopper has been working through the code removals for puppet 4. Not much else to say about that...

Data

I was hoping to have some data from our profiling information request…but we still don't have any :(

The data trap

--

Andrew Parker

an...@puppetlabs.com

Freenode: zaphod42

Twitter: @aparker42

Software Developer

Join us at PuppetConf 2014, September 20-24 in San Francisco - www.puppetconf.com

Kylo Ginsberg

unread,

Oct 9, 2014, 4:56:12 PM10/9/14

to puppe...@googlegroups.com

On Thu, Oct 9, 2014 at 12:24 PM, Andy Parker <an...@puppetlabs.com> wrote:

** Next PR Triage Wednesday, October 15th @ 10:00 am Pacific. **

Priorities
Puppet 3.7.2
CFacter on the march
New puppet doc implementation
Code removal for puppet 4
Commentary
Puppet 3.7.2 is in full swing. The big unknown for me was the memory leak (PUP-3345) that had been reported, but we hadn't reproduced. Well, it's been reproduced. It happens when directory environments are in use and environment_timeout != unlimited. Whenever it refreshes the environment cache it seems to leak memory. This is pretty critical for us since it will be in the next PE version, work on which is taking a lot of our time at the moment.

CFacter is continuing. Solaris support is still in the works and so is the toolchain for building it on all the various platforms. I believe you can see the build scripts at https://github.com/puppetlabs/cfacter-build

Clarification: that repo has build support only for Solaris and Windows. For new-enough Linux and OS X the build instructions are trivially short and are in the readme: https://github.com/puppetlabs/cfacter/blob/master/README.md

Kylo

Puppet-dev conversations of note:
* PUP-1073 - How to allow installing packages with the same name in different packaging systems in a single catalog?
This thread had been going on for a long time and had died off in March without a conclusion. The issue was brought back to my attention this past week and so I started the conversation up again with another proposal. There are some very valid concerns about what impact loosening the constraints will have on catalogs and flapping resources; however, it appears that the decision is to go ahead with the latest proposal. I still need to reply to the thread to conclude it and update tickets, etc, etc. I'm going to target a fix a 4.0.
* Behavior of apply + ENC
Felix presented some options about how command line arguments should behave wrt to environments + enc + command line overrides. There was a lot of agreement about a "command line args win" solution, but RI feels that "ENC wins" is the better solution. I'm waiting with baited breath for the conclusion :)
* Removing types/providers from core
Ah yes. This old thread. But it's going somewhere this time! Daenny and Kylo submitted a PR for removing nagios from core at the contributor summit and now Daenny is following up with working out a plant for removing the rest. I'd really like to see this drive to a conclusion that we take action on this time. At the moment a lot of things hinge on getting packaging setup to allow pulling in modules so that they don't live in the core repo, but are available in a standard puppet installation.

Puppet strings 0.1.0 has been released! Hailee put a lot of work into this to get it ready and Charlie did a lot of great work to show that it can be done. The next steps are for people to try it out and send us feedback/fixes.

Whopper has been working through the code removals for puppet 4. Not much else to say about that...

Data

I was hoping to have some data from our profiling information request…but we still don't have any :(

The data trap

--
Andrew Parker
an...@puppetlabs.com
Freenode: zaphod42
Twitter: @aparker42
Software Developer

Join us at PuppetConf 2014, September 20-24 in San Francisco - www.puppetconf.com

--
You received this message because you are subscribed to the Google Groups "Puppet Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to puppet-dev+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/puppet-dev/CANhgQXtSbWs9O%2Bimdn14vXddxyNTCdQTN%2B0mDZ%2BiJXgcePb2wQ%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

--

Kylo Ginsberg | ky...@puppetlabs.com | irc: kylo | twitter: @kylog

Felix Frank

unread,

Oct 9, 2014, 8:40:05 PM10/9/14

to puppe...@googlegroups.com

On 10/09/2014 09:24 PM, Andy Parker wrote:
> I'm waiting with baited breath for the conclusion :)

Then I hope you're in Portland these days, so that someone can check
your vital signs occasionally ;-)

Pete Brown

unread,

Oct 9, 2014, 9:40:15 PM10/9/14

to puppe...@googlegroups.com

On 10 October 2014 06:56, Kylo Ginsberg <ky...@puppetlabs.com> wrote:

On Thu, Oct 9, 2014 at 12:24 PM, Andy Parker <an...@puppetlabs.com> wrote:
** Next PR Triage Wednesday, October 15th @ 10:00 am Pacific. **

Priorities
Puppet 3.7.2
CFacter on the march
New puppet doc implementation
Code removal for puppet 4
Commentary
Puppet 3.7.2 is in full swing. The big unknown for me was the memory leak (PUP-3345) that had been reported, but we hadn't reproduced. Well, it's been reproduced. It happens when directory environments are in use and environment_timeout != unlimited. Whenever it refreshes the environment cache it seems to leak memory. This is pretty critical for us since it will be in the next PE version, work on which is taking a lot of our time at the moment.

CFacter is continuing. Solaris support is still in the works and so is the toolchain for building it on all the various platforms. I believe you can see the build scripts at https://github.com/puppetlabs/cfacter-build

Clarification: that repo has build support only for Solaris and Windows. For new-enough Linux and OS X the build instructions are trivially short and are in the readme: https://github.com/puppetlabs/cfacter/blob/master/README.md

Hi Gang,

Any idea when cfacter will show up in the nightly repos?

I am keen to add support for it (and the nightly repos) to my puppet management module.

Pete.

To view this discussion on the web visit https://groups.google.com/d/msgid/puppet-dev/CALsUZFHrFSV0Ayz9AD8VVOKYY9%3DT1MspvCRBKBzzuLnVuAnt%3Dg%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

--

Pete Brown

Director and Primary Systems Engineer

Abstract IT Pty Ltd.

Bruce Downs

unread,

Oct 13, 2014, 12:16:16 PM10/13/14

to puppe...@googlegroups.com

On Thursday, October 9, 2014 2:56:12 PM UTC-6, Kylo Ginsberg wrote:

On Thu, Oct 9, 2014 at 12:24 PM, Andy Parker <an...@puppetlabs.com> wrote:
** Next PR Triage Wednesday, October 15th @ 10:00 am Pacific. **

Priorities
Puppet 3.7.2
CFacter on the march
New puppet doc implementation
Code removal for puppet 4
Commentary
Puppet 3.7.2 is in full swing. The big unknown for me was the memory leak (PUP-3345) that had been reported, but we hadn't reproduced. Well, it's been reproduced. It happens when directory environments are in use and environment_timeout != unlimited. Whenever it refreshes the environment cache it seems to leak memory. This is pretty critical for us since it will be in the next PE version, work on which is taking a lot of our time at the moment.

CFacter is continuing. Solaris support is still in the works and so is the toolchain for building it on all the various platforms. I believe you can see the build scripts at https://github.com/puppetlabs/cfacter-build

Clarification: that repo has build support only for Solaris and Windows. For new-enough Linux and OS X the build instructions are trivially short and are in the readme: https://github.com/puppetlabs/cfacter/blob/master/README.md

Are these builds publicly consumable? On windows, I get a chocolatey error retrieving http://pl-build-tools.delivery.puppetlabs.net/Windows_NT/x86_64-4.8.2-release-win32-seh-rt_v3-rev4.7z. Did I miss a step?

i.e.

PS C:\Users\bdowns\Documents\cfacter-build> make build

C:/ProgramData/chocolatey/lib/make.3.81.4/tools/bin/make -e toolchain

make[1]: Entering directory `C:/Users/bdowns/Documents/cfacter-build'

c:\windows\system32\windowspowershell\v1.0\powershell.exe .\bin\Windows_NT\wget.ps1 http://pl-build-tools.delivery.puppetlabs.net/Windows_NT/x86_64-4.8.2-release-win32-seh-rt_v3-rev4.7z

Exception calling "DownloadFile" with "2" argument(s): "The remote server returned an error: (503) Server Unavailable."

At C:\Users\bdowns\Documents\cfacter-build\bin\Windows_NT\wget.ps1:3 char:1

+ (New-Object net.webclient).DownloadFile($args[0], $dest)

+ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

+ CategoryInfo : NotSpecified: (:) [], ParentContainsErrorRecordException

+ FullyQualifiedErrorId : WebException

make[1]: *** [fetched/x86_64-4.8.2-release-win32-seh-rt_v3-rev4.7z] Error 1

make[1]: Leaving directory `C:/Users/bdowns/Documents/cfacter-build'

make: *** [build] Error 2

Michael Smith

unread,

Oct 13, 2014, 1:14:31 PM10/13/14

to puppe...@googlegroups.com

The cfacter-build project hasn't been designed as a public build chain; it refers to internal resources for source and pre-compiled packages. You can try to fix the Makefiles to work externally, but it's probably easier to follow the build instructions at https://github.com/puppetlabs/cfacter/blob/master/WINDOWS.md.

To view this discussion on the web visit https://groups.google.com/d/msgid/puppet-dev/4f503290-264b-4981-aec0-07b2c2540b3b%40googlegroups.com.

Kylo Ginsberg

unread,

Oct 15, 2014, 12:36:38 PM10/15/14

to puppe...@googlegroups.com

On Thu, Oct 9, 2014 at 6:40 PM, Pete Brown <pe...@abstractit.com.au> wrote:

On 10 October 2014 06:56, Kylo Ginsberg <ky...@puppetlabs.com> wrote:
On Thu, Oct 9, 2014 at 12:24 PM, Andy Parker <an...@puppetlabs.com> wrote:
** Next PR Triage Wednesday, October 15th @ 10:00 am Pacific. **

Priorities
Puppet 3.7.2
CFacter on the march
New puppet doc implementation
Code removal for puppet 4
Commentary
Puppet 3.7.2 is in full swing. The big unknown for me was the memory leak (PUP-3345) that had been reported, but we hadn't reproduced. Well, it's been reproduced. It happens when directory environments are in use and environment_timeout != unlimited. Whenever it refreshes the environment cache it seems to leak memory. This is pretty critical for us since it will be in the next PE version, work on which is taking a lot of our time at the moment.

CFacter is continuing. Solaris support is still in the works and so is the toolchain for building it on all the various platforms. I believe you can see the build scripts at https://github.com/puppetlabs/cfacter-build

Clarification: that repo has build support only for Solaris and Windows. For new-enough Linux and OS X the build instructions are trivially short and are in the readme: https://github.com/puppetlabs/cfacter/blob/master/README.md

Hi Gang,

Any idea when cfacter will show up in the nightly repos?
I am keen to add support for it (and the nightly repos) to my puppet management module.

Hi Pete,

We just had a confab about this. It looks like we'll have cfacter in the nightlies starting in a week or two. We'll start with nightlies for just deb7/rhel7, and then add more distros/OSs as we build out our toolchains.

We'll definitely be posting announcements here as we go, so stay tuned!

Best,

Kylo

To view this discussion on the web visit https://groups.google.com/d/msgid/puppet-dev/CAM8R_x_UdCrDt6q8%3DJ7urJe3uOtLcpZVevdX7JSAOEwJ_Dw%2BYQ%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

Reply all

Reply to author

Forward