1

I have set up a cluster with ipyparallel on my local machine and wish to convert it to a dask cluster. My primary motivation for this is to be able to use the web ui with dask (via bokeh) to monitor the performance.

When I convert the ipyparallel cluster to a dask cluster, it does not start the bokeh web ui. Does anyone know how to launch the web UI either in the become_dask() function or after I have started dask

client_ipp = ipp.Client(profile='ipp_1')
dview  = client_ipp[:]
bview = client_ipp.load_balanced_view()

e = client_ipp.become_dask()
e

Client

Scheduler: tcp://XX.XXX.X.XX:XXXXX

Cluster

Workers: 8
Cores: 8
Memory: 33.60 GB

If I start a cluster directly with Dask I get:

from distributed import client
c3 = Client()
c3

Client

Scheduler: tcp://XXX.X.X.X:XXXXX
Dashboard: http://XXX.X.X.X:XXXXX

Cluster

Workers: 8
Cores: 8
Memory: 33.60 GB
Improve this question
2
  • 1
    This is discussed here: github.com/dask/distributed/issues/1268 . If you feel like fixing this that would be welcome. Commented Nov 29, 2017 at 13:16
  • Thanks for the info MRocklin. I will take a look... Commented Nov 30, 2017 at 17:30
Related questions
5
How to install dill in IPython?
1
python package distributed, get the address of the web UI with easy setup
12
How to enable ipyparallel for Jupyter?
Load 7 more related questions Show fewer related questions

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.