That looks to be the case, I'm going to try and login to the builders to see and also check how to specify a particular buildslave for the pre-cq.

Yes, that test build clearly ran on a GCE instance which won't work with VM Tests enabled. I'm going to try and track down how the trybot waterfall decides which builds go where.

Has the trybot waterfall been restarted since the lakitu config changes went in?

Reading through the trybot logic, it should do the right thing because it examines the build configs to decide what goes where. However, it only re-reads config changes when the waterfall is restarted.

I've restarted the waterfall restart that I think will fix this. Blocking this bug on the restart bug.

In the Chrome Infra codebase, see:
  build/masters/master.chromiumos.tryserver/chromiumos_tryserver_util.py.

This is the important bit of code:
  precq_builders = set(
      v['_template'] or k for k, v in configs.iteritems() if v.IsPreCqBuilder())
  precq_novmtest_builders = set(
      v['_template'] or k for k, v in configs.iteritems()
      if v.IsPreCqBuilder() and not v.HasVmTests() and not v.HasHwTests())

Thanks for looking into this.

lakitu changes were done in CL:344288 and CL:*258963. I don't know if waterfall was restarted since then.

Any idea of ETA for this?

Hopefully, this now fixed. Can someone run a tryjob to confirm?

Retrying committing https://chrome-internal-review.googlesource.com/#/c/259660

The latest pre-cq run once again got scheduled on "cros-standard18-c2" machine: https://uberchromegw.corp.google.com/i/chromiumos.tryserver/builders/pre-cq/builds/31073

Probably going to fail again?

Yep. I'm asking questions on the blocking bug. Maybe the restart hasn't really happened, only been scheduled. Or maybe the chromite ping wasn't bumped up before the restart.

The pin wasn't bumped (yes, confusing terminology). Nodir@ will bump it and restart that waterfall again. After that's done (and it may take a while), I expect this to be fixed.

The trybot waterfall was pin bumped and restarted, and it looks like the CL above passed PreCQ testing.

I think this is fixed!

Thanks! The pre-cq seems to be fixed.
Though it seems the lakitu_next-release builder started failing https://uberchromegw.corp.google.com/i/chromeos/builders/lakitu_next-release and the reason is unclear. Actually, only the 'cbuildbot' step is marked as failed, but the image build and test and all other steps are successful.
Not sure if its related to the restart.

I'm certain those issues are 100% unrelated, but looking.

Closing... please feel free to reopen if its not fixed.