John Ramirez
05/07/2020, 8:10 PMDaskCloudProviderEnvironment
and I keep getting this error
botocore.exceptions.NoRegionError: You must specify a region.
distributed.deploy.spec - WARNING - Cluster closed without starting up
any ideas?Laura Lorenz (she/her)
05/07/2020, 8:18 PMregion_name
kwarg to the DaskCloudProviderEnvironemnt
(which should get passed all the way down to where I think this is probably blowing up for you, at https://github.com/dask/dask-cloudprovider/blob/master/dask_cloudprovider/providers/aws/ecs.py#L652)
I think if you use another AWS configuring tool, like aws configure
from awscli
or writing up your own boto config, my guess is that boto should pull the default region from that for you insteadJohn Ramirez
05/07/2020, 8:22 PMregion_name
as an argumentJoe Schmid
05/07/2020, 8:26 PMJohn Ramirez
05/07/2020, 8:29 PMJoe Schmid
05/07/2020, 8:30 PMJohn Ramirez
05/07/2020, 8:36 PMFile "/Library/Frameworks/Python.framework/Versions/3.7/lib/python3.7/site-packages/distributed/deploy/cluster.py", line 73, in _close
await self._watch_worker_status_comm.close()
AttributeError: 'FargateCluster' object has no attribute '_watch_worker_status_comm'
Laura Lorenz (she/her)
05/07/2020, 9:16 PMdask.distributed
code, I’ll poke around but any chance you are trying to shutdown particularly quickly after starting, it looks like that attribute itself is `await`ed during cluster startJoe Schmid
05/07/2020, 9:20 PMdistributed
package? If it's easy, post the code you're using to create the cluster (removing any credential or sensitive info) and at what step in the process you see that error. I'm interested in helping to debug this!DaskCloudProviderEnvironment
.John Ramirez
05/08/2020, 1:08 AMfrom dask_cloudprovider import FargateCluster
cluster = FargateCluster(
image="daskdev/dask",
task_role_arn="arn:aws:iam:::role/ecsTaskExecutionRole",
execution_role_arn="arn:aws:iam:::role/ecsTaskExecutionRole",
# n_workers=1,
# scheduler_cpu=256,
# scheduler_mem=512,
# worker_cpu=256,
# worker_mem=512,
# scheduler_timeout="15 minutes",
)
Joe Schmid
05/08/2020, 1:19 PMprefecthq/prefect:latest
but the dask one should be fine for testing. (If you’d prefer we could debug this via DMs and then update the community channel with the results.)John Ramirez
05/08/2020, 1:20 PMaiobotocore==1.0.4
aiohttp==3.6.2
aioitertools==0.7.0
async-timeout==3.0.1
attrs==19.3.0
botocore==1.15.32
certifi==2020.4.5.1
chardet==3.0.4
click==7.1.2
cloudpickle==1.3.0
croniter==0.3.31
dask==2.15.0
dask-cloudprovider==0.2.0
distributed==2.15.2
docker==4.2.0
docutils==0.15.2
fsspec==0.7.3
HeapDict==1.0.1
idna==2.9
jmespath==0.9.5
locket==0.2.0
lz4==3.0.2
marshmallow==3.5.1
marshmallow-oneofschema==2.0.1
msgpack==1.0.0
multidict==4.7.5
mypy-extensions==0.4.3
numpy==1.18.2
pandas==1.0.3
pandas-market-calendars==1.2
partd==1.1.0
pendulum==2.0.5
prefect==0.10.7
psutil==5.7.0
pyarrow==0.17.0
python-box==4.2.3
python-dateutil==2.8.1
python-slugify==4.0.0
pytz==2019.3
pytzdata==2019.3
PyYAML==5.3.1
requests==2.23.0
ruamel.yaml==0.16.10
ruamel.yaml.clib==0.2.0
s3fs==0.4.2
six==1.14.0
sortedcontainers==2.1.0
tabulate==0.8.7
tblib==1.6.0
text-unidecode==1.3
toml==0.10.0
toolz==0.10.0
tornado==6.0.4
typing-extensions==3.7.4.2
urllib3==1.25.9
websocket-client==0.57.0
wrapt==1.12.1
yamlordereddictloader==0.4.0
yarl==1.4.2
zict==2.0.0
Joe Schmid
05/08/2020, 1:39 PMJohn Ramirez
05/08/2020, 1:40 PMRuntimeError: Scheduler failed to start
distributed.deploy.spec - WARNING - Cluster closed without starting up
Error in atexit._run_exitfuncs:
Joe Schmid
05/08/2020, 2:40 PMLuke Orland
08/25/2020, 7:09 PMJoe Schmid
08/25/2020, 7:14 PMLuke Orland
08/25/2020, 7:23 PMJoe Schmid
08/25/2020, 7:26 PM