[
https://issues.apache.org/jira/browse/JCLOUDS-671?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14097850#comment-14097850
]
Andrew Gaul commented on JCLOUDS-671:
-------------------------------------
[~arturo.fagundo] jclouds support lacks support Azure multi-part uploads with
non-repeatable `InputStream` payloads, similar to the Swift issue at
JCLOUDS-356. Your payload is not repeatable; the underlying
`S3Object.getInputStream` streams from the S3 blobstore and cannot rewind or
seek. You can work around this by copying the stream to disk:
{{noformat}}
File file = ...;
Files.asByteSink(file).writeFrom(S3Object.getObjectContent());
Payload payload = new ByteSourcePayloadPayload(Files.asByteSource(file));
{{noformat}}
If you must use a streaming `InputStream` payload jclouds you will need to add
support to jclouds. Can you look at the Swift support and write the Azure
equivalent?
> PutBlob call to Azure with Large Blob fails using InputStreamSupplierPayload
> ----------------------------------------------------------------------------
>
> Key: JCLOUDS-671
> URL: https://issues.apache.org/jira/browse/JCLOUDS-671
> Project: jclouds
> Issue Type: Bug
> Components: jclouds-blobstore
> Affects Versions: 1.7.2
> Environment: Uploading from my Windows 7 Eclipse environment to Azure
> Reporter: Arturo Fagundo
> Labels: azure
>
--
This message was sent by Atlassian JIRA
(v6.2#6252)
