Fix dhcpv6 counter test failure

[jcaiMR]

Description of PR

Summary:
Fixes # (issue)

Type of change

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

Back port request

• 201911
• 202012
• 202205

Approach

What is the motivation for this PR?

Fix test dhcpv6 counter failure
Dual tor issue:

1. Solicit counter not increase, random select dut will select standby port dut
2. Advertise counter not increase, downlink mac is not right, should always dut_mac

Single Tor issue:
Decline/Release/Information request/unknow counter not increase
test case send 14 kinds of packets almost at the same time.
As it's multicast packets, each packet get duplicated on fanout switch (maybe due to some topo setting), on one of local DUT,
observed 23 times duplicated packets (around 270 pakcets) send in 22 ms.
By default raw socket recv buffer size is 32k (270x128B = 34.5K), so some kinds of packets (always the packets at the last sequence like Release, Decline, Information request ...) get drooped and cause test case failure.

How did you do it?

1. toggle_all_simulator_ports_to_upper_tor and always use upper tor to avoid mux status inconsistent issue
2. use dut_mac for downlink interface mac (always same as port channel mac)
3. Add 1 second delay during testing for each kind of packets, in case any multicast packet duplicate issue happened

How did you verify/test it?

Test on single tor and dual tor DUT.

Any platform specific information?

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

Documentation

[lgtm-com]

This pull request fixes 1 alert when merging 4769461797b76f5d4141a1aa782de02851ac0e36 into c05753c - view on LGTM.com

fixed alerts:

• 1 for Unused import

---

Heads-up: LGTM.com's PR analysis will be disabled on the 5th of December, and LGTM.com will be shut down ⏻ completely on the 16th of December 2022. Please enable GitHub code scanning, which uses the same CodeQL engine ⚙️ that powers LGTM.com. For more information, please check out our post on the GitHub blog.

[lgtm-com]

This pull request fixes 1 alert when merging 90c632145d841b5f9ab15f7b5b6b03c14e363eba into fec265e - view on LGTM.com

fixed alerts:

• 1 for Unused import

---

Heads-up: LGTM.com's PR analysis will be disabled on the 5th of December, and LGTM.com will be shut down ⏻ completely on the 16th of December 2022. Please enable GitHub code scanning, which uses the same CodeQL engine ⚙️ that powers LGTM.com. For more information, please check out our post on the GitHub blog.

[StormLiangMS]

Approved with a minor suggestion.

[StormLiangMS]

is that 300 seconds delay too long? maybe 60 seconds do the job? @jcaiMR

[jcaiMR]

Sure, 60 should enough here, will apply in diff.

[StormLiangMS]

LGTM

[wangxin]

@jcaiMR Cherry-picking this PR to 202012 branch has conflicts that are hard to resolve. Can you create separate PR to 202012 branch to have this change?