Skip to content

AttributeError in handle_missing_dep #4199

@fjetter

Description

@fjetter

I can see a lot of logged errors for an AttributeError in the handle_missing_dep coroutine, e.g. in https://travis-ci.org/github/dask/distributed/jobs/739654791

Note: There are attribute errors for WorkerState and TaskState

It seems there is no test failure because of it but these exceptions seem to be retried and I could imagine this to be the cause why I can see many tests currently timing out w/out reason

This was probably introduced with the TaskState in #4107

tornado.application - ERROR - Exception in callback functools.partial(<bound method IOLoop._discard_future_result of <tornado.platform.asyncio.AsyncIOLoop object at 0x7ff0de6dfa90>>, <Task finished name='Task-96079' coro=<Worker.gather_dep() done, defined at /home/travis/build/dask/distributed/distributed/worker.py:1960> exception=CommClosedError()>)
2278Traceback (most recent call last):
2279  File "/home/travis/build/dask/distributed/distributed/worker.py", line 1980, in gather_dep
2280    self.validate_state()
2281  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2912, in validate_state
2282    self.validate_task(ts)
2283  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2867, in validate_task
2284    self.validate_task_flight(ts)
2285  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2854, in validate_task_flight
2286    assert ts.key in self.in_flight_workers[ts.coming_from]
2287AssertionError
2288
2289During handling of the above exception, another exception occurred:
2290
2291Traceback (most recent call last):
2292  File "/home/travis/miniconda/envs/dask-distributed/lib/python3.8/site-packages/tornado/ioloop.py", line 743, in _run_callback
2293    ret = callback()
2294  File "/home/travis/miniconda/envs/dask-distributed/lib/python3.8/site-packages/tornado/ioloop.py", line 767, in _discard_future_result
2295    future.result()
2296  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2082, in gather_dep
2297    self.batched_stream.send(
2298  File "/home/travis/build/dask/distributed/distributed/batched.py", line 146, in send
2299    raise CommClosedError
2300distributed.comm.core.CommClosedError
2301distributed.worker - ERROR - Handle missing dep failed, retrying
2302Traceback (most recent call last):
2303  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2124, in handle_missing_dep
2304    self.bad_dep(dep)
2305  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2105, in bad_dep
2306    "Could not find dependent %s.  Check worker logs" % str(dep.ts)
2307AttributeError: 'TaskState' object has no attribute 'ts'
2308distributed.worker - ERROR - Handle missing dep failed, retrying
2309Traceback (most recent call last):
2310  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2124, in handle_missing_dep
2311    self.bad_dep(dep)
2312  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2105, in bad_dep
2313    "Could not find dependent %s.  Check worker logs" % str(dep.ts)
2314AttributeError: 'TaskState' object has no attribute 'ts'
2315
2316

and


distributed.worker - ERROR - Handle missing dep failed, retrying
2309Traceback (most recent call last):
2310  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2124, in handle_missing_dep
2311    self.bad_dep(dep)
2312  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2105, in bad_dep
2313    "Could not find dependent %s.  Check worker logs" % str(dep.ts)
2314AttributeError: 'TaskState' object has no attribute 'ts'
2315
2316During handling of the above exception, another exception occurred:
2317
2318Traceback (most recent call last):
2319  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2117, in handle_missing_dep
2320    deps = {dep for dep in deps if dep.dependents}
2321  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2117, in <setcomp>
2322    deps = {dep for dep in deps if dep.dependents}
2323AttributeError: 'Worker' object has no attribute 'dependents'
2324distributed.worker - ERROR - Handle missing dep failed, retrying
2325Traceback (most recent call last):
2326  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2124, in handle_missing_dep
2327    self.bad_dep(dep)
2328  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2105, in bad_dep
2329    "Could not find dependent %s.  Check worker logs" % str(dep.ts)
2330AttributeError: 'TaskState' object has no attribute 'ts'
2331
2332During handling of the above exception, another exception occurred:
2333
2334Traceback (most recent call last):
2335  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2117, in handle_missing_dep
2336    deps = {dep for dep in deps if dep.dependents}
2337  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2117, in <setcomp>
2338    deps = {dep for dep in deps if dep.dependents}
2339AttributeError: 'Worker' object has no attribute 'dependents'
2340
2341During handling of the above exception, another exception occurred:
2342
2343Traceback (most recent call last):
2344  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2117, in handle_missing_dep
2345    deps = {dep for dep in deps if dep.dependents}
2346  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2117, in <setcomp>
2347    deps = {dep for dep in deps if dep.dependents}
2348AttributeError: 'Worker' object has no attribute 'dependents'
2349distributed.worker - ERROR - Handle missing dep failed, retrying
2350Traceback (most recent call last):
2351  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2124, in handle_missing_dep
2352    self.bad_dep(dep)
2353  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2105, in bad_dep
2354    "Could not find dependent %s.  Check worker logs" % str(dep.ts)
2355AttributeError: 'TaskState' object has no attribute 'ts'
2356
2357During handling of the above exception, another exception occurred:
2358
2359Traceback (most recent call last):
2360  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2117, in handle_missing_dep
2361    deps = {dep for dep in deps if dep.dependents}
2362  File "/home/travis/build/dask/distributed/distributed/worker.py", line 2117, in <setcomp>
2363    deps = {dep for dep in deps if dep.dependents}
2364AttributeError: 'Worker' object has no attribute 'dependents'
2365
2366During handling of the above exception, another exception occurred:

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions