tronweb

Tron comes with a simple but functional web-based UI to tronview. You can use tronweb to view your jobs and drill down into individual job runs, actions and action output. Currently, tronweb.py lives in the web/ directory of the tron source distribution.

tronweb requires the Tornado library.

Installation

You can run tronweb directly from the tron source distribution. It will require the tron package to be installed, and depends on Tornado.

Running tronweb

Before you run tronweb, you’ll need to set up a config.py file in the tronweb root directory. An example config file is included, config.py.example. The config file only includes a few attributes, most of which are self-explanatory:

config={
    'port': 8888,
    'trond_url':'http://localhost:8089',
    'output_url': "http://localhost:8889/{job}/{run}/{action}",
}

port defines the port tronweb runs on to handle requests. trond_url is the URL that trond listens on for requests. You can generally leave these two fields alone.

output_url is used by tronweb to build the URL used for downloading job output. tronweb does not serve these files itself because they may be quite large. You may wish to configure Apache, nginx, or some other web server to serve them.

job, run, and action will be replaced by the job name, job run name, and action name, which map directly to the path on disk where trond stores output files. For example, given run 10 of the job ‘call_center_report’ with a single action ‘report’, the url mapping above will generate stdout and stderr links of:

http://localhost:8889/call_center_report/call_center_report.10/call_center_report.10.report.stdout
http://localhost:8889/call_center_report/call_center_report.10/call_center_report.10.report.stderr

If you want a simple solution for serving output files, you can use Python’s built-in HTTP server:

$ cd working/directory/for/tron
$ python -m SimpleHTTPServer 8889

Once you have config.py set up, you simply start the server by running it:

$ python tronweb.py