What happened:
When starting a Dask distributed.SSHCluster I cannot access the Dask dashboard. The server returns 404: Not Found.
However, the dashboard is accessible when starting a local cluster via client = distributed.Client().
What you expected to happen:
I'd expect the Dask dashboard to show up.
Minimal Complete Verifiable Example:
python3 -m venv .venv
source .venv/bin/activate
python -m pip install dask distributed bokeh jupyter-server-proxy asyncssh
import distributed
cluster = distributed.SSHCluster(["127.0.0.1", "127.0.0.1"], remote_python='YOUR_DIRECTORY/.venv/bin/python')
client = distributed.Client(cluster.scheduler_address)
print(client.scheduler_info()['services']) # Output: {'dashboard': 8787}Anything else we need to know?:
pip freeze output:
aiohttp==3.8.1
aiosignal==1.2.0
anyio==3.6.1
argon2-cffi==21.3.0
argon2-cffi-bindings==21.2.0
async-timeout==4.0.2
asyncssh==2.10.1
attrs==21.4.0
beautifulsoup4==4.11.1
bleach==5.0.0
bokeh==2.4.3
cffi==1.15.0
charset-normalizer==2.0.12
click==8.1.3
cloudpickle==2.1.0
cryptography==37.0.2
dask==2022.5.0
defusedxml==0.7.1
distributed==2022.5.0
entrypoints==0.4
fastjsonschema==2.15.3
frozenlist==1.3.0
fsspec==2022.5.0
HeapDict==1.0.1
idna==3.3
Jinja2==3.1.2
jsonschema==4.5.1
jupyter-client==7.3.1
jupyter-core==4.10.0
jupyter-server==1.17.0
jupyter-server-proxy==3.2.1
jupyterlab-pygments==0.2.2
locket==1.0.0
MarkupSafe==2.1.1
mistune==0.8.4
msgpack==1.0.3
multidict==6.0.2
nbclient==0.6.3
nbconvert==6.5.0
nbformat==5.4.0
nest-asyncio==1.5.5
numpy==1.22.4
packaging==21.3
pandocfilters==1.5.0
partd==1.2.0
Pillow==9.1.1
prometheus-client==0.14.1
psutil==5.9.1
ptyprocess==0.7.0
pycparser==2.21
Pygments==2.12.0
pyparsing==3.0.9
pyrsistent==0.18.1
python-dateutil==2.8.2
PyYAML==6.0
pyzmq==23.0.0
Send2Trash==1.8.0
simpervisor==0.4
six==1.16.0
sniffio==1.2.0
sortedcontainers==2.4.0
soupsieve==2.3.2.post1
tblib==1.7.0
terminado==0.15.0
tinycss2==1.1.1
toolz==0.11.2
tornado==6.1
traitlets==5.2.1.post0
typing_extensions==4.2.0
urllib3==1.26.9
webencodings==0.5.1
websocket-client==1.3.2
yarl==1.7.2
zict==2.2.0
Environment:
- Dask version: 2022.05.0
- Python version: 3.10.4
- Operating System: Ubuntu 22.04 LTS
- Install method (conda, pip, source): pip
What happened:
When starting a Dask
distributed.SSHClusterI cannot access the Dask dashboard. The server returns404: Not Found.However, the dashboard is accessible when starting a local cluster via
client = distributed.Client().What you expected to happen:
I'd expect the Dask dashboard to show up.
Minimal Complete Verifiable Example:
python3 -m venv .venv source .venv/bin/activate python -m pip install dask distributed bokeh jupyter-server-proxy asyncsshAnything else we need to know?:
pip freezeoutput:Environment: