No offset option in method "openInputStream()" of "com.microsoft.azure.storage.blob.CloudBlockBlock" instance

[helderribeirosousa]

Hello,

I'm developing an application that leverages Azure's Blob Storage API for Java.
Everything went fine till i found the need to read a CloudBlob starting from a given offset.
Sure i can start reading from 0 and discard the bytes till i reach the desired offset, but this is both a major waste of network resources and impractical for large files (>300MB) with an average download speed from Azure to my server of 2.5MBps.

I need something like BlobInputStream openInputStream(long offset).

Since other blob storage providers support this, am i missing on something or is this feature not supported by Azure Storage SDK?

[rickle-msft]

Hello!

Thank you for your patience. I think the skip method on a BlobInputStream should do the trick? Let us know if this is insufficient. http://azure.github.io/azure-storage-java/com/microsoft/azure/storage/blob/BlobInputStream.html#skip-long-

[helderribeirosousa]

Thank you @rickle-msft.
I thought that method was inherited from InputStream and wasn't optimised for network usage. Looks like i should have searched for the documentation you have provided.
Still, please consider the BlobInputStream openInputStream(long offset) overload...

[rickle-msft]

This feature has been added in the latest release of the v8 sdk. The constructor for BlobInputStream now has an overload which takes an offset and a count.

I will close this issue now as the feature has been added, but please feel free to post again if you have further requests.