Adds subnet_id to launch method.

[MatthewRalston]

This pull request fixes two issues with boto.

• Fixes an issue with inter-node communication permissions. It seems that boto's authorize method still requires the ip_protocol, source port (0) and destination port (65535) along with the src_group parameter.
• Boto has an issue with specifying a subnet and a groupName simultaneously. The solution described by this thread seems to be specifying the subnet_id's instead of the subnet name.

Let me know what you want to see here.

1. When I pass a subnet_id to boto's run_instances method, I get the following error. This seems to be a incompletely addressed idiosyncracy of the boto library and how it prefers for subnets to be specified. See this stackoverflow post and this github issue related to the run_instances call with a subnet_id. It seems that boto prefers subnet ids over subnet names in this method.

[2015-10-26 10:10:34,542: DEBUG/MainProcess] <?xml version="1.0" encoding="UTF-8"?>
<Response><Errors><Error><Code>InvalidParameterCombination</Code><Message>The parameter groupName cannot be used with the parameter subnet</Message></Error></Errors><RequestID>e72077dd-f31d-4c94-b0ac-970f6c41fe56</RequestID></Response>
[2015-10-26 10:10:34,542: ERROR/MainProcess] 400 Bad Request
[2015-10-26 10:10:34,542: ERROR/MainProcess] <?xml version="1.0" encoding="UTF-8"?>
<Response><Errors><Error><Code>InvalidParameterCombination</Code><Message>The parameter groupName cannot be used with the parameter subnet</Message></Error></Errors><RequestID>e72077dd-f31d-4c94-b0ac-970f6c41fe56</RequestID></Response>
[2015-10-26 10:10:34,543: ERROR/MainProcess] Problem launching an instance: The parameter groupName cannot be used with the parameter subnet (code InvalidParameterCombination; status 400)
Traceback (most recent call last):
  File "/home/ralstonm/Projects/bioblend/bioblend/cloudman/launch.py", line 177, in launch
    placement=placement)
  File "/home/ralstonm/.pyenv/versions/cloudlaunch/lib/python2.7/site-packages/boto/ec2/connection.py", line 973, in run_instances
    verb='POST')
  File "/home/ralstonm/.pyenv/versions/cloudlaunch/lib/python2.7/site-packages/boto/connection.py", line 1208, in get_object
    raise self.ResponseError(response.status, response.reason, body)
EC2ResponseError: EC2ResponseError: 400 Bad Request
<?xml version="1.0" encoding="UTF-8"?>
<Response><Errors><Error><Code>InvalidParameterCombination</Code><Message>The parameter groupName cannot be used with the parameter subnet</Message></Error></Errors><RequestID>e72077dd-f31d-4c94-b0ac-970f6c41fe56</RequestID></Response>

[MatthewRalston]

The travis build seems to have failed for a reason unrelated to my changes. I can build this locally:

~/Projects/bioblend >pip install -e `pwd`
DEPRECATION: --download-cache has been deprecated and will be removed in the future. Pip now automatically uses and configures its cache.
Obtaining file:///home/ralstonm/Projects/bioblend
Requirement already satisfied (use --upgrade to upgrade): requests>=2.4.3 in /home/ralstonm/.pyenv/versions/cloudlaunch/lib/python2.7/site-packages (from bioblend==0.7.0.dev0)
Requirement already satisfied (use --upgrade to upgrade): requests-toolbelt in /home/ralstonm/.pyenv/versions/cloudlaunch/lib/python2.7/site-packages (from bioblend==0.7.0.dev0)
Requirement already satisfied (use --upgrade to upgrade): boto>=2.9.7 in /home/ralstonm/.pyenv/versions/cloudlaunch/lib/python2.7/site-packages (from bioblend==0.7.0.dev0)
Requirement already satisfied (use --upgrade to upgrade): pyyaml in /home/ralstonm/.pyenv/versions/cloudlaunch/lib/python2.7/site-packages (from bioblend==0.7.0.dev0)
Requirement already satisfied (use --upgrade to upgrade): six in /home/ralstonm/.pyenv/versions/cloudlaunch/lib/python2.7/site-packages (from bioblend==0.7.0.dev0)
Installing collected packages: bioblend
  Running setup.py develop for bioblend
Successfully installed bioblend-0.7.0.dev0

Some of the idiosyncracies of the boto call have been described where I am using them in Cloudlaunch

[nsoranzo]

creaInvalid... explicitlyting: I think you pasted in the wrong place!

[nsoranzo]

The TravisCI failed on Galaxy dev branch due to Galaxy-related problems, but on the other Galaxy branches there are 3 flake8 issues introduced by your PR:

./bioblend/cloudman/launch.py:113:58: E231 missing whitespace after ','
./bioblend/cloudman/launch.py:176:46: E265 block comment should start with '# '
./bioblend/cloudman/launch.py:308:50: E231 missing whitespace after ','

[afgane]

Tested this manually and it worked fine so merged in (with a minor whitespace fix).

[nsoranzo]

I've fixed the TravisCI failures due to the Galaxy eggs-to-wheels migration, future PRs should be fine.

[MatthewRalston]

Thank you all