An application I'm writing uses S3, and unfortunately my client can't switch for a variety of reasons. I really really (really) want to use nirvanix, but I won't be able to use it as our primary data store. I heard there is an S3 conversion tool, but the conversion tool isn't enough (it's actually too much), as I want to get objects from S3, and put them in nirvanix to transcode on an object by object basis. The scalability problem for me is that I absolutely positively need the tell nirvanix to get the data from S3 without me downloading the data first. I'm thinking of using the semi-public url scheme (urls that expire a few minutes after they are issued) that S3 provides with it's query string auth tokens. (ridiculously psuedoish code) This is how I'm seeing the current flow of data into nirvanix: data = read_from_public_url_into_memory('http://some_url') node = nirvanix.create_node('file.txt') node.put(data) What I want to do is something like this: node = nirvanix.create_node('file.txt'') node.put_data_from_public_url('http://some_url') Possible? Am I missing something in the docs?

I think you are looking for this: 

 http://developer.nirvanix.com/sitefiles/1000/API.html#_ToSideload

 btw, I also missed it the first time. The name "Sideload" is really self-explanatory...

I tried it and it worked great... 

ciao Luca

 

Awesome, thanks!

I also use Amazon webservices. There is another way to do mass upload to your nirvanix account

Now i dont know how well you know the ec2 AMI structure. But you could load out your s3 content

within the /mnt/ directory of your AMI and use the nirvanix nput tool to upload your content to nirvanix.

The nput app uses curl so there might be some delay but your files are transfered completely.

And without any issues.

 

I have used the x.large AMI to upload lots of HD video to my nirvanix account.