search

guillermo-carrasco / bcbio-nextgen-monitor

Monitor the status of bcbio-nextgen analyses
MIT License
4 stars 2 forks source link

Deal with analyses restarts #39

Closed guillermo-carrasco closed 8 years ago

guillermo-carrasco commented 8 years ago

If a user restarts a failed analyses without removing the old log file, bcbio will continue writing on the same log file without cleaning it. This is not a bug but a design choice, since previous run information may be useful.

The problem though is that the monitor can't deal with this yet, and it will find "repeated" steps and mess it up. Possible solutions are:

  1. When finding a stacktrace, just continue reading instead of stopping. Annotate somewhere that a stacktrace was found.
  2. Implement the logic to detect this situation, by for example detecting 2 "Organize samples" steps. Then show somehow to the user (create two tables/flowcharts instead of just one?)

If both scenarios the monitor needs to deal with repeated step id's.

Thanks @lpantano and @ohofmann for pointing this out. Suggestions on how to deal with this are very much welcome :)

guillermo-carrasco commented 8 years ago

This seems to be a nice way:

screen shot 2015-11-20 at 15 46 28

Just add a tab for every re-run, so that you can easily see how many re-runs there have been, where did they fail, etc.

lpantano commented 8 years ago

oh that's so sweet!!!!!

guillermo-carrasco commented 8 years ago

:smiley: glad you like it @lpantano !