How big a factor is blob size on retrieval time in Azure Storage?

Question

I am storing text files in Azure Blob Storage. The files will be on the order of 1MB, but I could theoretically reduce that size by perhaps 30% at the cost of significantly increasing my application logic complexity.

I'm leaning toward just using the simpler but larger files. However, I wanted to know what factor blob size would have on retrieval time. Is it negligible or could there be a significant difference? I'm retrieving from a web server directly within the same datacenter as the blobs.

Also, is any compression automatically applied to blobs being sent within the datacenter? (As text files with lots of repeated content, they would compress very well.)

Thanks for any pointers!

Answer 1

It depends on your E2E usage scenario but for simply uploading text files to Azure Blob Storage I would suggest going with the simpler but larger ~1MB files as the difference after reducing the size will probably be negligible.

You can also take a look at the Azure Storage Scalability and Performance Targets - https://azure.microsoft.com/en-us/documentation/articles/storage-scalability-targets/

Answer 2

Considering the network speed with in the data centre, there would not be any issues with 1MB text files.Or atleast they would be negligible.

And also as you don't want to increase the app's complexity to handle this data, better even don't explicitly try to do any compressed file transfers as the size 1MB is ok to be transferred with out any compression. And I doubt that you would end up doing some logic to decompress the received compressed file.

Ref this thread for Compression details - Does Windows Azure Blob Storage support serving compressed files similar to Amazon S3?