Azure Ultra Disk

2019: Start of service

In August, 2019 Microsoft started a cloud service for storage of the Big Data Azure Ultra Disk which works at a basis of solid state drives.

The company calls a new solution the distributed block storage system of new generation intended for Azyre virtual machines which need high capacity and fast data transmission in both directions.

One of servers based on which cloud services of Microsoft Azure function; a demonstration sample at the OCP Summit 2019 conference

It is said in the statement of Microsoft that Azure Ultra Disk will become good option for "the applications using Big Data, such as SAP HANA, the best DBMS and applications with a huge amount of transactions".

Several options of data storage — from 4 Gbytes to 64 Tbyte are available to users of Azure Ultra Disk. Clients can also select capacity from range from 300 Mbps to 2000 MB/s, and transactions of input-output can be set ranging from 1200 up to  160,000 units per second.

The technical director of Microsoft Azure Mark Russinovich wrote in the blog that users can dynamically scale the capacity of the used drives according to the requirements to workloads without need to reboot the virtual machines. Settings can be changed on the fly, and they will become effective less than in an hour, he noted.

The representative of the company also explained that Azure Ultra Disk uses locally excess storage for storage of three copies of data in one zone of availability Azure (Azure Availability Zone).

Besides, the technology of "the client of a virtual disk" which provides metadata for logical display of a disk is provided. Using these metadata the client can interact directly with data storage servers that reduces delays and provides performance at the level of corporate a flash storages, Russinovich reported.^[1]

Notes