python - Dask分布式诊断网页不起作用

原文 标签 python parallel-processing distributed-computing dask

Dask Distributed Diagnostic Webpage not working

I've gotten dask up and running on my cluster, but I can't seem to access the diagnostic webpage. The landing page is visible, see below:

enter image description here

But all the links just hang and never load the page.

The scheduler started fine with this output:

[hoffmand@h05u06 ~]$ dask-scheduler --scheduler-file dask-scheduler.json
distributed.scheduler - INFO - -----------------------------------------------
distributed.scheduler - INFO -   Scheduler at:   tcp://10.36.105.16:8786
distributed.scheduler - INFO -       bokeh at:              0.0.0.0:8788
distributed.scheduler - INFO -        http at:              0.0.0.0:9786
distributed.bokeh.application - INFO - Web UI: http://127.0.0.1:8787/status/
distributed.scheduler - INFO - -----------------------------------------------
distributed.scheduler - INFO - Register tcp://10.36.107.15:37780
distributed.scheduler - INFO - Starting worker compute stream, tcp://10.36.107.15:37780
Answer

I just experienced the same issue, and apparently, with bokeh <= 0.12.5 and tornado = 4.5 (see https://github.com/dask/distributed/issues/1027)

So either:

  • Downgrade your tornado to 4.4.3
  • Upgrade bokeh to 0.12.7 (or at least 0.12.6 apparently)

I just upgraded to bokeh 0.12.7 and it works like a charm.

翻译

我已经在群集上启动dask并在其中运行,但似乎无法访问诊断网页。着陆页可见,请参见下文:

enter image description here

但是所有链接只是挂起,永远不会加载页面。

调度程序开始时显示以下输出:

[hoffmand@h05u06 ~]$ dask-scheduler --scheduler-file dask-scheduler.json
distributed.scheduler - INFO - -----------------------------------------------
distributed.scheduler - INFO -   Scheduler at:   tcp://10.36.105.16:8786
distributed.scheduler - INFO -       bokeh at:              0.0.0.0:8788
distributed.scheduler - INFO -        http at:              0.0.0.0:9786
distributed.bokeh.application - INFO - Web UI: http://127.0.0.1:8787/status/
distributed.scheduler - INFO - -----------------------------------------------
distributed.scheduler - INFO - Register tcp://10.36.107.15:37780
distributed.scheduler - INFO - Starting worker compute stream, tcp://10.36.107.15:37780
最佳答案
我刚刚遇到了同样的问题,显然,bokeh <= 0.12.5和tornado = 4.5(请参见https://github.com/dask/distributed/issues/1027)

所以:


将您的tornado降级为4.4.3
bokeh升级到0.12.7(或至少0.12.6 apparently)


我刚刚升级到bokeh 0.12.7,它就像一个魅力。
相关推荐

python - KeyError读取csv文件并传输到数组

python - Doc2Vec句子聚类

python - Scrapy python csv输出的每一行之间都有空白行

python - 如何检测填充了哪些圆圈+ OpenCV + Python

python - 是否可以更改当地人的命令?

python - 为Python2和Python3编写Unicode正则表达式

python - pip安装tesserocr失败,并显示错误“ tesserocr的建筑轮失败”

python - webbrowser python在同一选项卡中打开新的URL。这行得通吗?

python - 使用Python进行轨迹聚类/聚合

python - 使用model.fit_generator时keras val非常慢