Just a quick one here, been pulling my hair out for a while about this.
Basically I was getting timeouts when trying to copy data from my web role compute instance and the blob storage.
Couldn’t find any reason or proper answer.
Then found out that I happened to setup my compute in Anywhere Europe, and my blob to Western Europe. So they were in different data centres, which is obvious when you think about it. But just didn’t think. So I recreated my compute instance in Western Europe, and throughput has improved.