ptftests: fail advanced-reboot fast on unreachable neighbors

[Ryangwaite]

Description of PR

Summary:
Fixes # (issue)
MSFT ADO: 37767011

When a cEOS/SONiC neighbor is unreachable from the PTF (e.g. broken container networking), advanced-reboot would hang for ~11 minutes and report the misleading failure "DUT hasn't shutdown in 600 seconds".

Root cause: peer_state_check SSH threads wedged inside paramiko.connect on its ~130s default timeout while wait_until_cpu_port_down blocked on their Queue.put('cpu_down'), so the 600s control-plane-down timer expired before any SSH error surfaced.

To avoid, this there is now a preliminary check that we can access the neighbours over ssh and if not, abort the test early with a more visible reason for failing.

Testing

Without this change

These logs are seen:

2026-05-01 00:28:48 : Error in HostDevice: Traceback (most recent call last):
  File "/root/ptftests/py3/advanced-reboot.py", line 1768, in peer_state_check
    self.fails[ip], self.info[ip], self.cli_info[ip], self.logs_info[ip], self.lacp_pdu_times[ip] = ssh.run()
                                                                                                    ^^^^^^^^^
  File "/root/ptftests/py3/arista.py", line 142, in run
    self.connect()
  File "/root/ptftests/py3/arista.py", line 52, in connect
    self.conn.connect(self.ip, username=self.login,
  File "/root/env-python3/lib/python3.11/site-packages/paramiko/client.py", line 384, in connect
    sock.connect(addr)
TimeoutError: [Errno 110] Connection timed out
...
2026-05-01 00:37:29 : Reachability watcher - checking VLAN GW IP
2026-05-01 00:37:30 : DUT hasn't shutdown in 600 seconds: Traceback (most recent call last):
  File "/root/ptftests/py3/advanced-reboot.py", line 427, in timeout
    res = async_res.get(timeout=seconds)
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/lib/python3.11/multiprocessing/pool.py", line 770, in get
    raise TimeoutError
multiprocessing.context.TimeoutError

2026-05-01 00:37:30 : ==================================================

Ran 1 test in 678.613s

That 600s timeout is misleading when the true error is that the ceos neighbours couldn't be reached.

With this change

======================================================================
ERROR: advanced-reboot.ReloadTest
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/root/ptftests/py3/advanced-reboot.py", line 1475, in runTest
    self._verify_neighbors_reachable()
  File "/root/ptftests/py3/advanced-reboot.py", line 1472, in _verify_neighbors_reachable
    raise RuntimeError(msg)
RuntimeError: Neighbor SSH unreachable from PTF: [('172.16.131.32', 'TimeoutError: timed out'), ('172.16.131.33', 'TimeoutError: timed out'), ('172.16.131.34', 'TimeoutError: timed out'), ('172.16.131.35', 'TimeoutError: timed out')]

----------------------------------------------------------------------
Ran 1 test in 87.427s

FAILED (errors=1)

This is a saving of ~10min and it has a clearer error for debugging.

Type of change

• Bug fix
• Testbed and Framework(new/improvement)
• New Test case
  • Skipped for non-supported platforms
• Test case improvement

Back port request

• 202205
• 202305
• 202311
• 202405
• 202411
• 202505
• 202511

Approach

What is the motivation for this PR?

Save testbed resources and improve debuggability

How did you do it?

Added a preliminary neighbour check before starting the main part of the test.

How did you verify/test it?

Happy and sad path testing with the fix in place.

Any platform specific information?

N/A

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

N/A

Documentation

N/A

[mssonicbld]

/azp run

[azure-pipelines]

Azure Pipelines successfully started running 1 pipeline(s).