As ever more applications are built to be cloud native, services like Redis are essential to enable these applications to perform and scale under heavy loads. To allow developers to utilize the speed and flexibility of Redis without the hassle of operation, Azure offers Azure Cache for Redis, a fully managed Redis service. A third-party benchmark recently showed that Azure Cache for Redis can improve latency and throughput performance by up to 800% when added to existing Azure architectures. To build on this powerful technology, we announced an innovative collaboration with Redis Labs almost a year ago to bring their Redis Enterprise software to Azure as a part of Azure Cache for Redis. Customers both large and small have been excited to try the expanded capabilities during the preview, and developers have found novel ways to solve their application challenges with the offering. Today, we’re proud to announce the general availability of the Enterprise and Enterprise Flash tiers of Azure Cache for Redis.
What’s in the box?
The new Enterprise offerings enhance and expand the range of use cases for Redis today. Redis modules from Redis Labs, like RediSearch and RedisTimeSeries, enable new use-cases for Redis such as real-time search and IoT data ingestion. At the same time, the offering also improves your existing Redis experience by making caches larger, more resilient, and capable of being distributed around the globe. The powerful features available include:
Active Geo-Replication (in Preview)
Higher availability—architected for 99.99% with zone redundancy, and 99.999% with active geo-replication.
Redis Modules, including
The latest version of Redis—Redis 6.0
Redis on Flash, using both DRAM and NVMe storage to enable cache sizes up to 13TB at a lower price per GB.
If you currently use Redis Enterprise on-prem, this offering allows you to enjoy many of your favorite features in a native Azure solution. Streamlined billing, familiar security and monitoring tools, and the ability to use your Azure spend commitment are all included. Even better, support for the Enterprise offering on Azure is backed all day, every day by the Microsoft support team, who have a direct line to the experts at Redis Labs if additional assistance is needed.
New in the Enterprise offering is a much-requested upgrade: the ability to geo-replicate data across caches in an active-active configuration. Now in preview, this feature allows data to be written locally in more than one region, with changes automatically replicated to and merged with other regions. Active geo-replication enables you to create a worldwide distributed cache with local latency access for users around the globe. This feature opens many exciting possibilities:
One key application is handling user sessions across regions. A huge advantage of Redis is the blazing-fast latencies gained from accessing data in-memory. If a user is far away from the data center holding the cache, however, this advantage is limited due to network latency. With active geo-replication, user session data can follow a user without complicated data management on the backend. The data is simply duplicated to the Redis instance in the closest data center, with all changes synchronized automatically.
Another important use case is with applications that are distributed worldwide. Active geo-replication greatly simplifies the data layer by reducing the need to manage geographies. For instance, a mobile game might have a worldwide leaderboard. Instead of developing infrastructure to share and rank scores between regions, high scores from each user can simply be written to a Redis sorted set that is distributed across regions worldwide. Through conflict-free resolution in active geo-replication, scores will be ranked seamlessly.
One significant advantage of active geo-replication is that it makes a Redis deployment substantially more resilient to single-region failures. This, combined with the option to deploy Redis across multiple availability zones in a single region, provides two additional high-availability options. Deploying across availability zones gives up to 99.99% availability, while active geo-replication is designed for up to 99.999% availability—less than six minutes of downtime per year. This increased resiliency is essential to many retail and financial services customers who rely on Redis for inventory or pricing systems.
Redis Modules + Redis 6.0
The Redis Enterprise offering on Azure doesn’t just boost geographic distribution and availability, however. It also includes powerful features that enable new use cases. With the introduction of Redis 6.0, Azure Cache for Redis users gain access to the Redis Streams data type for the first time. Streams can be used to implement messaging systems, allowing Kafka-like functionality where clients can intelligently consume items from the stream. Three Redis Modules are additionally supported: RedisBloom, RediSearch, and RedisTimeSeries:
Customers like SitePro have already started using these highly-effective modules in their workflows, for example using RedisTimeSeries for IoT data ingestion:
"We are extremely excited about the added functionality in the Azure Cache for Redis, Enterprise Tiers. We know, love, and have used Redis for years as a key component in our industry leading real-time IoT control & monitoring platform. The Enterprise tiers allow us to do more with Redis by leveraging the RedisTimeSeries module to support data collection and analytics from hundreds of thousands of IoT sensors. "