In case you missed it, here are the Key Microsoft Azure updates in December, 2016
Larger Block Blobs in Azure Storage
What is this: Azure Storage now supports Block Blobs of up to 4.77 TB in size. This is in comparison to the previous limit of 195 GB. Also increased…the maximum allowable block size from 4 MB to 100 MB, while keeping the limit on the number of blocks committed to a single Blob constant at 50,000. Range GETs continue to be supported on larger Block Blobs. Customers can immediately begin taking advantage of this improvement in any existing Blob Storage or General Purpose Storage Account across all Azure regions. It applies to the latest versions of the SDK and AZCopy command-line utility.
Azure Feature Pack for SSIS (SQL Server 2016 Update)
What is this: Lots of great updates to SSIS and Azure services integration. For example, Azure Data Lake Store for both source/destination and other big data scenarios are included and available for orchestration by SSIS control flow.
Microsoft R Server v9.x
What is this: Microsoft R Server 9.x release. This includes the MicrosoftML commercial R package (for Windows and SQL Server 2016) which includes new “FAST” algorithms that have implicit parallelization and can scale well beyond memory. It also includes better support for Spark as well improved operationalization options. The virtual machine with MRS v9.x is also available on Azure.
Azure Service Updates (new clean/consolidated look)
What is this: The Azure service updates site has a nice new consolidated list view look. Great place to check out the too numerous Azure updates list too long to list all here.
Other Noteworthy Learning in December 2016
Azure Data Lake Store – Performance of Copying Between Blob Storage
What is this: Great video that demonstrates some performance metrics of copying between Azure Blob Storage and Azure Data Lake Store. This is a great reference for customers who are hesitant to move to ADLS and are worried about having to move persisted files between two Azure cloud storage options. Furthermore, certain popular Azure features (i.e. SQL Data Warehouse PolyBase) don’t support ADLS yet.
Note: The demonstrated performance using Azure Data Factory to move files between blob storage and ADLS is about 1TB copied per 10 minutes; pretty quick! Therefore, you can ingest about 30 Petabytes of data per day. (demo starts at 24-minute mark). Demo at the 34-minute mark that shows an Azure Data Lake Store seeded with 5.02 PB of data. Demo at the 36.5-minute mark that shows a single file with a size of 1 PB.
SQL Server 2016 SP1 (Matrix of functionality capabilities)
What is this: Since we rolled out some Enterprise SKU features into Standard and Web SKUs, this post outlines their new limitations. Very useful for SQL Server on Azure VM sizing conversations.
What is this: Over 600 bloggers that post R news and tutorials. Our own Microsoft PG team makes quite a few posts here regarding advanced analytics on R. Highly recommended for your customers and/or yourself to keep up with the latest on R (including MSFT R announcements).
AzureSMR R Package
What is this: This is an R package that allows data scientists to manage selected Azure resources within an R session. This essentially allows information workers to not only craft machine learning scripts, but also provision the required big data/analytics Azure cloud infrastructure.