Timeout while exporting model (On-Premise)

Machine Learning
#1

MODEL_NAME = “iris_knn”
model.export(“Resources/sklearn/iris”, MODEL_NAME, metrics={‘accuracy’: acc})

Polling irisFlowerClassifier version 1 for model availability.
get model:/hopsworks-api/api/project/1163/models/irisFlowerClassifier_1?filter_by=endpoint_id:1163
Model not available during polling, set a higher value for synchronous_timeout to wait longer.

Getting this error in demo project. Don’t know what am I doing wrong?

#2

Hi FeSS,

The model export is relying on a message bus - epipe to propagate all models metadata so that they are available for searching. Can you check if the epipe service is up and running?
What version of hopsworks did you install?
In order to check the status of the service you can ssh into the vm/machine and then do systemctl status epipe and try to restart it systemctl restart epipe.

Regards,
Alex

1 Like
#3

@FeSS If you are administrator in Hopsworks you can also monitor and restart services from the Admin UI. You can access it from by clicking on your profile at the top-right corner, then Admin, Services.

#4

Thank you. Restarting epipe helped alot :slight_smile: