I came back to Japan.

This is Akiyama, an engineer.

I participated in Microsoft Build 2017 , which was held in Seattle on 5/10~12 local time, and this time I am posting a report after returning to Japan.

You can watch the Build on Channel9 .

Big Data on Azure Blob Storage

In this article, we'll summarize the session titled Around Azure Storage. This is because on the third day, the last day of this session, there were no keynotes, and when I was listening carelessly, there was more announcement than I expected, so I hurriedly wrote this article.

You can check the video of the session at the following URL.

Big data workloads with Azure Blob Storage

Click here for the Day 1 keynote, click here for the Day 2 keynote , and click here for a summary of Azure in general Please refer to.

Storage

Most of them are announcements related to storage itself.

AzCopy

AzCopy, a utility tool that handles storage data through the CLI, was previously only available on Windows, but the Linux version, AzCopy on Linux, is now in preview.
For the link, please refer to the Channel9 video at the beginning of the 35-minute period.

Data Protection

Featured support for data storage for blobs was announced. This feature is optional and can protect blobs for up to 365 days and up to 10 generations. It is scheduled for GA from July to September 2017.

Tiered Storage

Blobs have Hot for high-frequency access and Cool for low-frequency access.
Storage has been announced that allows you to switch between three states per blob, including an inaccessible Archive. The preview will be announced soon.

Is it better to use Archive as a storage for old data that is rarely opened? It seems that the usage will vary depending on how much you are charged for the switching operation.

Encryption

The encryption feature for writing data to blobs has only been enabled and disabled until now, but it seems that you will be able to manage the encryption key when you enable it. It is scheduled to be released in preview in the second half of 2017.

Encryption In Transit

Storage was available for HTTP/HTTPS access, but an option was announced to limit it to HTTPS only. It is scheduled to be released in the second half of 2017.

Network Access Control

The ability to restrict access to Storage to VirtualNetwork or IP addresses has been announced. It is scheduled to be released in preview in the second half of 2017.

AAD Authentication and RBAC for Blob Storage

We've announced the ability to control access with IAM in Azure on a per-container basis. It is scheduled to be released in preview in the second half of 2017.

Storage Account Scalability

Previously, 500TB was the storage account capacity limit, but it seems that the cap will be above 5PB. This also results in 100k+ requests/sec and 100+ Gbps IO. It is scheduled to be released in preview in the summer of 2017.

Virtual Machine

There was also a little announcement around storage for Virtual Machines.

VM Storage Performance & Scale

It has been announced that the maximum disk size will be 4TB, which was capped at 1TB for both Standard and Premium. Also, it is unclear whether the upper limit of each spec applies to Standard or Premium, but it seems that it will be up to 256TB, 80,000IOPS, and 2GB/s. (Previously, it was capped at 1TB, 5,000IOPS, and 200MB/s, respectively) and 32GB and 64GB are supported in Premium. It is in preview in the Midwest US region.

I'm happy to be able to use SSDs easily.

Conclusion

Since storage is not only a standalone service but also a technology that underpins other Azure services, the evolution of storage will also lead to the evolution of Azure as a whole.
As a user, I am grateful.