Platform Team Week of February 2nd, 2013

15 views
Skip to first unread message

Andy Parker

unread,
Feb 10, 2014, 12:34:25 PM2/10/14
to

Week of February 2nd, 2013

** Next PR Triage Wednesday, February 12th @ 9:30 am Pacific. **

Priorities



  2. PE 3.2 support/PE 3.3 development and planning

  3. Ability to list dynamic environments (PUP-536)

  4. Facter 2.0 implementation (PUP-552)
  5. Windows ACLs (PUP-574)

Commentary

Facter 1.7.5 continues. We are putting together an RC2 to get RHEL7 packages as part of it. Adrien is polishing up the Facter 2 work and fixing some bugs that were uncovered in the automated integration tests (puppet testing against facter)

The new environment system is continuing. After talking again to some people in PL who had asked for the feature, it became apparent that the feature was a bit off from what they needed. Thankfully it was just wrong in what the REST API was providing and so that should be a quick fix. However it also came out that the configurability (PUP-1596) is more urgently needed than I had thought.

Rob has started plowing ahead on Windows ACL support.

There are now benchmarks to track some performance characteristics of puppet. They live in the benchmarks directory in the repo and can be run with "rake benchmark:<name>:run". I ended up doing

Changes Moving Forward

This is more of a possible change. What times would work best for those who want to attend the PR triage? We chose 9am Pacific because it works well for us, who are mostly on the west coast, and it seemed like it might be ok for those in Europe. However, I've heard tell that this might not be the case. Suggestions for other times?
Data



Inline image 1


So….JIRA seems to have new charts….
Inline image 2
--
Andrew Parker
Freenode: zaphod42
Twitter: @aparker42
Software Developer

Join us at PuppetConf 2014September 23-24 in San Francisco - http://bit.ly/pupconf14
cd6d2766c7152877a554e0bf34e924e5.png
4f800cc753fbd27ca5198ba8285f6d8c.png

Andy Parker

unread,
Feb 10, 2014, 12:47:11 PM2/10/14
to
It has been brought to my attention that I speak in riddles of incomplete sentences. Clarification inline :)


On Mon, Feb 10, 2014 at 9:34 AM, Andy Parker <an...@puppetlabs.com> wrote:

Week of February 2nd, 2013

** Next PR Triage Wednesday, February 12th @ 9:30 am Pacific. **

Priorities



  2. PE 3.2 support/PE 3.3 development and planning

  3. Ability to list dynamic environments (PUP-536)

  4. Facter 2.0 implementation (PUP-552)
  5. Windows ACLs (PUP-574)

Commentary

Facter 1.7.5 continues. We are putting together an RC2 to get RHEL7 packages as part of it. Adrien is polishing up the Facter 2 work and fixing some bugs that were uncovered in the automated integration tests (puppet testing against facter)

The new environment system is continuing. After talking again to some people in PL who had asked for the feature, it became apparent that the feature was a bit off from what they needed. Thankfully it was just wrong in what the REST API was providing and so that should be a quick fix. However it also came out that the configurability (PUP-1596) is more urgently needed than I had thought.

Rob has started plowing ahead on Windows ACL support.

There are now benchmarks to track some performance characteristics of puppet. They live in the benchmarks directory in the repo and can be run with "rake benchmark:<name>:run". I ended up doing


.... something.

Oh yeah! I ended up following up on the benchmark with some profiling and tuning. Every benchmark scenario also has a profile to generate a kcachegrind calltrace output of running that scenario. This was really useful in tracking down some problems. In working on this I've already submitted a few pull requests to address some slowdowns that I found. They are linked from PUP-751, which ended up overlapping with Brice's work on PUP-1592. Brice's investigations are taking him to the agent side of the equation, whereas my work stayed entirely in compilation on the master.
 
Changes Moving Forward

This is more of a possible change. What times would work best for those who want to attend the PR triage? We chose 9am Pacific because it works well for us, who are mostly on the west coast, and it seemed like it might be ok for those in Europe. However, I've heard tell that this might not be the case. Suggestions for other times?
Data



Inline image 1


So….JIRA seems to have new charts….
Inline image 2
--
Andrew Parker
Freenode: zaphod42
Twitter: @aparker42
Software Developer

Join us at PuppetConf 2014September 23-24 in San Francisco - http://bit.ly/pupconf14
4f800cc753fbd27ca5198ba8285f6d8c.png
cd6d2766c7152877a554e0bf34e924e5.png

Erik Dalén

unread,
Feb 11, 2014, 3:29:05 PM2/11/14
to Puppet Developers
On 10 February 2014 18:34, Andy Parker <an...@puppetlabs.com> wrote:

Changes Moving Forward

This is more of a possible change. What times would work best for those who want to attend the PR triage? We chose 9am Pacific because it works well for us, who are mostly on the west coast, and it seemed like it might be ok for those in Europe. However, I've heard tell that this might not be the case. Suggestions for other times?
Data

For me personally wednesdays are tricky, but I don't think it will be possible to find a day that suits everyone. But perhaps you could do it some other day every second week or something?

I think it would also be better to have it an hour or two later. it is now 18:00-20:00 CET which is too early to get home and make dinner before, but also means you would have to stay quite late at work to participate in. if it started 19 or 20 it would be easier to comfortably get home and make dinner before it starts.
It should also make it easier for people on the east coast, now it is right in the middle of lunch, right?

my 2c

--
Erik Dalén
Reply all
Reply to author
Forward
0 new messages