Traceback (most recent call last):
    File "/b/s/w/ir/third_party/catapult/telemetry/telemetry/testing/serially_executed_browser_test_case.py", line 194, in <lambda>
      return lambda self: based_method(self, *args)
    File "/b/s/w/ir/content/test/gpu/gpu_tests/gpu_integration_test.py", line 76, in _RunGpuTest
      self.RunActualGpuTest(url, *args)
    File "/b/s/w/ir/content/test/gpu/gpu_tests/info_collection_test.py", line 62, in RunActualGpuTest
      (expected_vendor_id, detected_vendor_id))
  AssertionError: Vendor ID mismatch, expected 32902 but got 0


ccameron, can you take a look at this, as you're pixel wrangling next week.

Actually ccameron and I are swapping, so this is on my plate.

Just took a look, I'm not sure if anything can be done about it - it must be extremely flaky; doesn't seem to have failed even once in the past 200 builds.

If this symptom ever happens on the Mac Minis (basically, the "Mac Intel" GPU bots), it means the Raritan KVM got disconnected, and a Labs ticket (type "Build Infrastructure", component "Infra > Labs") needs to be filed to get the machine fixed.

In this case, that shard ran on build569-m4 (this is the "shard #0" link from the step):

https://chromium-swarm.appspot.com/task?id=36750ebd21d56c10&refresh=10&show_raw=1

Here's this bot's log:

https://chromium-swarm.appspot.com/bot?id=build569-m4&sort_stats=total%3Adesc

It's still failing to identify the GPU. I'm going to click the "shut down gracefully" button and file a dependent Labs ticket to fix the KVM and bring it back up.

BTW -- any Telemetry based test will show the GPU vendor and device IDs as 0x0 if this happens. It also looks like gl_tests (a native C++ executable) will complain that it failed to identify the GPU on at least some of its tests.

https://chromium-swarm.appspot.com/task?id=36839d2a9147a810&refresh=10

This hasn't flaked in a long time:
https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Mac%20Release%20%28Intel%29?limit=200

Closing as not reproducible.