In an increasingly global and online world, the speed, availability, and consistency of data has never been more important. We’ve seen customers of all sizes take advantage of the outstanding performance of Azure Cache for Redis, but we’ve also heard your requests for a Redis solution that offers even higher availability and more robust geographic consistency. That’s why we’re excited to announce the latest element of our collaboration with Redis Corporation—the general availability of active geo-replication in our Enterprise offering of Azure Cache for Redis.
Active geo-replication background
Active geo-replication is a powerful tool that enables Azure Cache for Redis clusters to be linked together for seamless active-active replication of data. In other words, you can write to one Redis cluster and your data will be automatically copied to the other linked clusters, and vice versa. Data is quickly duplicated with strong eventual consistency between clusters. This multi-primary, multi-write architecture is built on conflict-free replicated data types (CRDTs), a groundbreaking technology that enables seamless conflict resolution.
Use-cases for active geo-replication
Active geo-replication gives developers a powerful tool to tackle otherwise frustrating technical challenges, such as:
Enterprise-grade Availability
As Redis steadily increases in popularity and adoption, it is becoming an essential part of mission-critical applications. Many customers in the financial services, retail, and software industries need Redis in both high availability and disaster recovery scenarios. The Enterprise tiers of Azure Cache for Redis help deliver both. Zone redundancy already enables Enterprise Azure Cache for Redis to be resilient to zone-level outages and reach a 99.99% availability SLA. Adding active geo-replication helps further protect against region-level outages, boosting the availability SLA up to 99.999%.
Even better, experiencing the benefits of higher availability is straightforward. Even if there is an outage in one region, the caches in other regions will have the latest synchronized copy of the data in your cache. And when the region comes back online, the original cache will automatically be updated with data written to the linked regions during the outage.
Local Latency Performance
Customer expectations for speedy application performance are sky high. Nine in ten shoppers will abandon a retail website if it is too slow. As applications become global, this presents a challenge with caching. Your cache may be quick, but if your customers are talking to a cache that is halfway around the world, network latency can erase the benefits of the cache. With active geo-replication, you can get local latency performance while maintaining cache consistency because users and applications can be directed to the geo-replicated cache closest to them.
Global Synchronization of Data
Managing multiple sources of data can be a frustrating challenge. Scale or latency constraints can force you to silo your app infrastructure by geography, which then can introduce data consistency issues. Some applications are designed to have the same experience across borders, which confounds the problem even further. For example, a game developer might want all players to see a global leaderboard which ranks players across regions. While Redis is a common way to implement a leaderboard, aggregating multiple independent leaderboards from separate Redis instances is a hassle. With active geo-replication, however, the linked Redis instances can all update the same shared sorted set leaderboard, meaning a new high score in Tokyo will be automatically ranked next to a score from a player in London.
Try it on Azure
Azure Cache for Redis gives you a fully-managed Redis experience. With full portal integration, configuring active geo-replication is straightforward, and gives you a wide selection of regions to choose from in the ever-growing Azure footprint. Try active geo-replication today in the Enterprise and Enterprise flash tiers of Azure Cache for Redis, read how to configure the feature, or check out a simple active geo-replication demo.