We have the concept of an analysis runner that runs tasks for the website. E.g. false color index visualization for the website. The false color indices for the site represent a kind of mass analysis. Their output needs to be stored in a fashion that is not the same as an ordinary analysis task.
NOTE: this is not a custom/user/analysis jobs scenario.
We have to run the false-color generation over all of our website files.
We currently have a more general analysis worker. It contains a lot of good logic but for our system level analysis runner, some of that logic needs to be overridden / customised.
Current plan is to sub-class the current analysis runner and hardcode / customise the following:
the command to execute?
the output directory
the source of the config file taken care of by #12
Note these changes only affect the enqueue portions of the code.
Ideally the dequeuer (the part of the code that actually executed jobs) will remain the same.
We have the concept of an analysis runner that runs tasks for the website. E.g. false color index visualization for the website. The false color indices for the site represent a kind of mass analysis. Their output needs to be stored in a fashion that is not the same as an ordinary analysis task.
NOTE: this is not a custom/user/analysis jobs scenario.
We have to run the false-color generation over all of our website files.
We currently have a more general analysis worker. It contains a lot of good logic but for our system level analysis runner, some of that logic needs to be overridden / customised.
Current plan is to sub-class the current analysis runner and hardcode / customise the following:
the source of the config filetaken care of by #12Note these changes only affect the enqueue portions of the code.
Ideally the dequeuer (the part of the code that actually executed jobs) will remain the same.