dask-scheduler does not launch Bokeh
See original GitHub issueI am seeing this starting with distributed 1.21.8 and bokeh 1.0.1 but also with the very latest versions. Starting dask-scheduler will not start bokeh any more. We are using our own compiled Python 3.6.7 and install all packages via pip. I tested it with the latest Anaconda Python 3.7 and that works fine so it seems to be an issue with our Python install …
petersen@koshuf0:~$ dask-scheduler --port 12121 --bokeh-port 12122 --host $SLURMD_NODENAME --pid-file dask-scheduler.pid
distributed.scheduler - INFO - -----------------------------------------------
distributed.scheduler - INFO - Clear task state
distributed.scheduler - INFO - Scheduler at: tcp://172.30.1.100:12121
distributed.scheduler - INFO - Local Directory: /loc/scratch/27074823/scheduler-xu5904ps
distributed.scheduler - INFO - -----------------------------------------------
Connection to koshu-login closed by remote host.
below Scheduler at: I should be seeing
distributed.scheduler - INFO - bokeh at: 172.30.1.100:12122
but it is not there and bokeh is not started, there are no errors in the log. I would be grateful for any hint . There does not seem to be a --verbose or --debug option for dask-scheduler ?
Issue Analytics
- State:
- Created 5 years ago
- Comments:16 (15 by maintainers)
Top Results From Across the Web
Dask dashboard not starting when starting scheduler with api
Since bokeh is installed, I'd expect the dashboard to be started. When I start the scheduler using the command line however, the dashboard ......
Read more >Diagnostics (distributed) - Dask documentation
Bokeh must be installed in your scheduler's environment to run the dashboard. If it's not the dashboard page will instruct you to install...
Read more >Parallel processing with Dask scheduler - Guides - Guild AI
Unlike a queue, a Dask scheduler is capable of running operations ... for other runs to stop before starting staged runs (default is...
Read more >FAQ — arboreto 0.1.5 documentation - Read the Docs
Q: How can I use the Dask diagnostics (bokeh) dashboard? ... Bokeh error when launching Dask scheduler; Workers do not connect with Dask...
Read more >Cluster setup — pysmFISH 0 documentation - Linnarsson Lab
... use dask-distributed to manage not only the cluster scheduler/workers but also the parallel handling of the processing. The current pipeline can be...
Read more >Top Related Medium Post
No results found
Top Related StackOverflow Question
No results found
Troubleshoot Live Code
Lightrun enables developers to add logs, metrics and snapshots to live code - no restarts or redeploys required.
Start FreeTop Related Reddit Thread
No results found
Top Related Hackernoon Post
No results found
Top Related Tweet
No results found
Top Related Dev.to Post
No results found
Top Related Hashnode Post
No results found
Top GitHub Comments
Everything is fine. If you want the dashboard (which is useful) then you’ll have to install the additional library
bokeh
for visualizationor
I was getting this error - " Web dashboard not loaded. Unable to import bokeh". It’s not clear from reading the documentation on how to get more information on what’s causing this. I agree with the above that dask is impaired without the web dashboard as I don’t know if or how my workers are being utilized properly. Is there a verbose logging option I’ve missed?
I was able to fix it by going python3 -m pip install “dask[complete]” again. I guess I must have ran something which clobbered a version of something.