jython-processor-plugin-xxxx was available in a plugin that is not installed

Solved!
Shahrukh
Level 1
jython-processor-plugin-xxxx was available in a plugin that is not installed

Hi,

I was testing the dataiku test queries API of a prediction model which contains custom pre-processor API script.

The script is working fine and I'm able to train and score the model.

However, when i tried to run test query on prediction model API, i facing the below error.

jython-processor-plugin-xxxx was available in a plugin that is not installed. 

The plugin is already visible in installed section in admin page.

Is there something I'm missing here?

Shahrukh
0 Kudos
1 Solution
fchataigner2
Dataiker

if the step is meant to compute additional features, an option is of course to not make it a custom step but a regular PythonUDF step (ie use the code of the plugin step in a regular Python step, provided it runs as jython). Another possibility is to make a "python function endpoint" to get the raw features, apply the preprocessing, then call the api node as a "sub-service" (cf https://doc.dataiku.com/dss/latest/apinode/api/endpoints-api.html#calling-another-endpoint )

View solution in original post

3 Replies
fchataigner2
Dataiker

Hi,

if by "prediction model API" you refer to the API node or the test queries of the API services design section (which is actually an embedded API node), then it's expected, because API nodes don't have a notion of plugins, and thus cannot execute scripts containing custom steps.

0 Kudos
Shahrukh
Level 1
Author

Hi,

Thank you for the solution.

Can you please tell me what other options do I have here?

A custom python function api which creates those features and then fed to original api?

Shahrukh
0 Kudos
fchataigner2
Dataiker

if the step is meant to compute additional features, an option is of course to not make it a custom step but a regular PythonUDF step (ie use the code of the plugin step in a regular Python step, provided it runs as jython). Another possibility is to make a "python function endpoint" to get the raw features, apply the preprocessing, then call the api node as a "sub-service" (cf https://doc.dataiku.com/dss/latest/apinode/api/endpoints-api.html#calling-another-endpoint )