search

RevolutionAnalytics / AzureML

An R interface to AzureML(https://studio.azureml.net/) experiments, datasets, and web services.
Other
46 stars 22 forks source link

upload.dataset() can't deal with large data sets #97

Open fboylu opened 8 years ago

fboylu commented 8 years ago

while using upload.dataset() on a data frame, the following error is returned:

Error: AzureML returns error code: HTTP status code : 500 Maximum request length exceeded. Traceback:

  1. upload.dataset(labeledfeatures, ws, name = "labeledfeatures")
  2. try_fetch(url, handle = h)
  3. validate_response(r)
  4. stop(msg, call. = FALSE)

I was able to upload a smaller subset of the data frame without issues.

gilbertw commented 8 years ago

We need to switch to a chunked upload to handle the large payloads. See the python sdk change for reference, https://github.com/Azure/Azure-MachineLearning-ClientLibrary-Python/commit/4204e2b2f1540e6d9f936b19457048c5afdca860

stephlocke commented 7 years ago

Any ETA available on a resolution to this issue?

andrie commented 7 years ago

Pull requests are welcome. I don't think anybody is actively working on this issue.

stephlocke commented 7 years ago

Ok

There's no contributor guidelines and whatnot on this - before PRs can be accepted, does this need to be once-overed for Microsoft compliance stuff or is it ok as-is?

http://open.microsoft.com/