These are chat archives for ipython/ipython

Apr 2016
Steve Conover
Apr 03 2016 17:42 UTC
hi, I am running a jupyter notebook server and would like tot urn on debug level logging for the ipython kernel. I can't figure out a proper setting in I pretty much have the same problem as this person: Does anyone here have recommendations on how to do this?
I currently have "c.IPKernelApp.log_level = 'DEBUG'" in my config, and am running the server using "jupyter notebook --no-browser --debug". I see debug logging for jupyter, but not any spawned kernels.
(as an aside, what i would really like to do is get a jupyter server working in the pycharm debugger, and have the debugger properly attach to any kernel subprocesses. if anyone out there has gotten this to work, i would love to hear from you...)
Isaac Overcast
Apr 03 2016 23:57 UTC
Hi folks. First of all, I love ipython/ipyparallel, totally killer, thanks for all your work on it... Now, I have a question about ipyparallel. I'm using loadbalanced_view() and map_async() in an app that processes a _ton of data. I'm seeing a weird difference in memory usage between ipyparallel 4.0 (works great) and 5.0 (memory usage accumulates rapidly eventually exhausting system resources). This is for the exact same codebase for my app, just swapping out ipyparallel versions at build time. I'm wondering if this is something anyone has seen before.