Fine Grained ECMP test

[anish-n]

##Description of PR
This PR contains the changes for Fine Grained ECMP test.
It is based on the following specs:
https://github.com/Azure/SONiC/blob/master/doc/ecmp/fine_grained_next_hop_hld.md
https://github.com/Azure/SONiC/pull/623/files

Summary:

Type of change

• Bug fix
• Testbed and Framework(new/improvement)
• Test case(new/improvement)

Approach

This test creates Fine Grained ECMP objects on the DUT, and validates data plane for fine grained ecmp via PTF tests.

What is the motivation for this PR?

https://github.com/Azure/SONiC/blob/master/doc/ecmp/fine_grained_next_hop_hld.md

How did you do it?

Initially create a route with x next-hops, send 1000 flows, change the set of next-hops for all the possible fine grained ecmp scenarios and validate that the 1000 flows adjust as expected in the scenario

How did you verify/test it?

By running the PTF test

Any platform specific information?

Applicable to all platforms which support fine grained ecmp

Supported testbed topology if it's a new test case?

T0

Documentation

https://github.com/Azure/SONiC/blob/master/doc/ecmp/fine_grained_next_hop_hld.md

[lgtm-com]

This pull request introduces 11 alerts when merging 3214e86 into ca24f81 - view on LGTM.com

new alerts:

• 8 for Unused import
• 2 for Variable defined multiple times
• 1 for 'import *' may pollute namespace

[lgtm-com]

This pull request introduces 11 alerts when merging 87039c7 into 9cd2589 - view on LGTM.com

new alerts:

• 8 for Unused import
• 2 for Variable defined multiple times
• 1 for 'import *' may pollute namespace

[abdosi]

@anish-n minigraph based test cases will be added later

[anish-n]

@anish-n minigraph based test cases will be added later

Yes, will raise a separate PR to include minigraph based test case, this is because the test case for minigraph is expected to have a good deal of minigraph-related distinct changes and can be made into a separate PR. Also the minigraph changes are still open in PR form, so this would raise a dependency, so good to merge this now if possible, and have a 2nd PR to deal with the minigraph side.

[abdosi]

@nazariig Can you approve this ?