Add extended flag to CPU collector

[scottcunningham]

This allows us to produce both simple and complex metrics - this way, simple metrics can be aggregated, but we still have the option to drill-down into per-host metrics in situations that require them (eg. checking steal on an individual VM)

[coveralls]

Coverage remained the same at 59.729% when pulling 2bede02 on Ensighten:extended_cpu_metrics_upstream into ab59e12 on python-diamond:master.

[shortdudey123]

This should be documented and tests updated

[coveralls]

Coverage remained the same at 59.729% when pulling 85e92f3 on Ensighten:extended_cpu_metrics_upstream into ab59e12 on python-diamond:master.

[scottcunningham]

@shortdudey123 I've added docs (I ran make docs and then replaced the INSERT_EXAMPLES_HERE with the examples previously there), tried to start looking at tests but it seems like many tests are broken in master:

Ran 478 tests in 2.439s
FAILED (failures=2, errors=8, skipped=4)
make: *** [test] Error 1

They seem to be legitimate errors (eg. 'NoneType' object has no attribute 'CommandGenerator', False is not true etc), so it's probably not a problem with missing libraries or somesuch. Are there any docs on how the tests work or how to run a subset of the tests?

[shortdudey123]

Travis shows all tests passing currently so that means you are missing a needed library i your dev environment
Take a look at https://github.com/python-diamond/Diamond/blob/master/.travis.requirements.txt

[scottcunningham]

Thanks @shortdudey123, will give it a go tomorrow and hopefully return with some tests!

Are there any docs about this, or general dev setup, or a CONTRIBUTING.md with details about requirements for PRs? I couldn't find any when I looked but I might be missing something obvious. If not, I'm more than happy to contribute docs or instructions. Should I open a new issue for this?

[shortdudey123]

There really only is 1 doc file for development and it isn't too large:
https://github.com/python-diamond/Diamond/tree/master/docs/Development

A while back i added a vagrant box that does most of the setup and testing:
https://github.com/python-diamond/Diamond/blob/master/Vagrantfile#L78-L120

[coveralls]

Coverage decreased (-39.6%) to 20.084% when pulling c810a9a on Ensighten:extended_cpu_metrics_upstream into 8c7ea91 on python-diamond:master.

[coveralls]

Coverage decreased (-39.6%) to 20.084% when pulling c810a9a on Ensighten:extended_cpu_metrics_upstream into 8c7ea91 on python-diamond:master.

[scottcunningham]

Thanks for the info @shortdudey123. I've added tests for the extended flag, and also the simple flag (since I needed something to base the extended flag tests on and there were none).

Not sure why coveralls think the coverage has decreased so drastically, I'll take a look at that.

[kormoc]

Would it make more sense to add two new flags:

• aggregate defaulting to false
• individual defaulting to true

and having simple be an alias for aggregate true and individual false?