Answered by:
AWS and AZURE

Question
-
Answers
-
Hey Shalini,
The Reader module + Hive Query mode supports any publicly-facing Hadoop cluster using WebHDFS, not just HDInsight. The Reader can also ingress data from an unauthenticated HTTP address, if you can provide one from S3. However, there is no native S3 connector. In fact, data transfer speeds will be vastly improved by hosting your data on an Azure Storage account in the same datacenter as your Azure ML workspace, though I understand the desire to keep it in S3.
Regards,
AK
- Proposed as answer by neerajkh_MSFTModerator Thursday, September 10, 2015 4:08 AM
- Marked as answer by neerajkh_MSFTModerator Sunday, September 13, 2015 5:18 PM
All replies
-
Hi Shailini26,
I am not sure whether you would be able to connect the Machine learning contents from Amazon S3 to Azure.
https://aws.amazon.com/machine-learning/faqs/
As per the documentation, you cannot export the AML contents outside of Amazon. (AML-Amazon Machine Learning)
Connecting Amazon S3 Hadoop to Azure you can check the below docs for details.
http://dennyglee.com/2012/03/21/connecting-hadoop-on-azure-to-your-amazon-s3-blob-storage/
Hope this helps you.
Girish Prajwal
-
Hey Shalini,
The Reader module + Hive Query mode supports any publicly-facing Hadoop cluster using WebHDFS, not just HDInsight. The Reader can also ingress data from an unauthenticated HTTP address, if you can provide one from S3. However, there is no native S3 connector. In fact, data transfer speeds will be vastly improved by hosting your data on an Azure Storage account in the same datacenter as your Azure ML workspace, though I understand the desire to keep it in S3.
Regards,
AK
- Proposed as answer by neerajkh_MSFTModerator Thursday, September 10, 2015 4:08 AM
- Marked as answer by neerajkh_MSFTModerator Sunday, September 13, 2015 5:18 PM
-
-