[warm-reboot] Preboot sad path automation for n lag members

[neethajohn]

Signed-off-by: Neetha John nejo@microsoft.com

Description of PR

Covers cases where a single lag member or all lag members can be brought down from the neighbor end or DUT end before warm-reboot

Test case as follows

1. Bring down single/all lag members down from the neighbor or DUT end for the selected VMs as per the preboot_oper passed
2. Verify the BGP and LAG and lag member state for the selected VMs
3. Do a warm reboot
4. Verify the BGP and LAG and lag member state for the selected VMs and ensure they remain the same as preboot
5. Revert the preboot operation

Type of change

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

Approach

How did you do it?

• warm-reboot-multi-sad.yml
  • number of lag members down is appended with a ':' to the preboot oper type
    *.py
  • extended existing lag down case to bring down multiple interfaces
  • added necessary mapping to identify which lag members are down so that verification of lag member state can be done properly

How did you verify/test it?

Tested on T0 topology. Seeing some issues in the single lag member down case. Need investigation.

[pavel-shirshov]

Suggest to introduce one more method which uses change_neigh_lag_state().

The new method could work with many interface and use existing method to change one interface

[pavel-shirshov]

temp is not the best name in this case. I'd better use someting more meaningful

[neethajohn]

renamed

[yxieca]

It would be helpful to add an example of input string here for reference. Also the code has an implicit decision that only the last 2 values were used. You should add a comment to explain the choice.

[neethajohn]

added more comments