Overview of Memorystore for Redis

Memorystore for Redis provides a fully-managed service that is powered by the Redis in-memory data store to build application caches that provide sub-millisecond data access.

Memorystore for Redis offers several advantages over self-managed Redis:

  • Deploy what fits your needs. Memorystore for Redis allows you the flexibility to choose from different service tiers and sizes that fit your performance and operational needs. With a few clicks, you have the option to deploy a Basic Tier standalone Redis instance or a Standard Tier high availability Redis instance up to 300 GB.
  • Easily scale to get blazing speed. With Memorystore for Redis, you can easily achieve your latency and throughput targets by scaling up your Redis instances with minimal impact to your application's availability. Start with the lowest tier and smallest size, then grow your Redis instance as the needs of your application change.
  • Highly available and more secure. Redis instances are protected from the internet using private IPs and are further secured using Cloud Identity and Access Management role-based access control. Standard high availability instances are always replicated across zones and provide a 99.9% availability SLAs.
  • Focus on your application. Memorystore for Redis automates the complex operational tasks that are required to deploy and manage Redis. Tasks like provisioning, replication, failover, and monitoring are all automated. Applications connect to a single endpoint, eliminating any node-discovery logic and simplifying management and operations. Additionally, integration with Cloud Monitoring makes it easy to monitor your Redis instances.
  • Redis Protocol Compatible. Memorystore for Redis is fully Redis protocol compliant. You can move your applications using open source Redis to use Memorystore for Redis without any code changes. There is no need to learn new tools: all existing tools and client libraries just work.

What it's good for

Memorystore for Redis provides a fast, in-memory store for use cases that require fast, real-time processing of data. From simple caching use cases to real time analytics, Memorystore for Redis provides the performance you need.

  • Caching: Cache is an integral part of modern application architectures. Memorystore for Redis provides low latency access and high throughput for heavily accessed data, compared to accessing the data from a disk based backend store. Session management, frequently accessed queries, scripts, and pages are common examples of caching.

  • Gaming: Gaming is about capturing and keeping the user's attention. One key aspect that keeps users hooked on a game is the leaderboard. Everyone wants to see how they are progressing and where they stand. Making this experience snappy is critical, and with its in-memory store and data structure like Sorted Set, Memorystore for Redis makes it easy to maintain a sorted list of scores while providing uniqueness of elements. Player Profile is another piece of information that can be accessed frequently. Redis hash makes it fast and easy to store and access profile data.

  • Stream Processing: Whether processing a Twitter feed or stream of data from IoT devices, Memorystore for Redis is a perfect fit for streaming solutions. Combined with Dataflow, Memorystore for Redis provides a scalable, fast in-memory store for storing intermediate data that thousands of clients can access with very low latency.


The following table describes the features that Memorystore for Redis provides:

  • Fully managed: Deploying and maintaining a Redis instance can be time consuming. Memorystore for Redis provides patching, 24x7 threat monitoring, failure detection, and automatic failover, allowing you to spend more time on building your applications.
  • Simple deployment: Easily deploy a single instance or highly available Redis instance by choosing Basic Tier or Standard Tier from the Google Cloud Console or by automating deployment using Cloud SDK or Cloud Client Libraries.
  • High availability: Memorystore for Redis instances in the Standard Tier are replicated across zones, monitored for health and have fast automatic failover. Standard Tier instances also provide an SLA of 99.9%. For more information, see High Availability.
  • Enterprise-grade security: Redis instances are protected from the internet using private IPs, and access to instances is controlled and limited to applications running on the same Virtual Private Network as the Redis instance.

    Instances are further secured using Cloud IAM roles, which enables granular control over who can manage and access the Redis instance. Also, Memorystore for Redis is integrated with Cloud Audit Logging, enabling you to answer "who did what and when".

  • Scale, as needed: Memorystore for Redis enables scaling your instances up to a maximum of 300 GB and supports up to 12 Gbps of network throughput. With the ability to scale instance size seamlessly, you can start small and increase the size of the instance as needed. When scaling a Standard Tier instance, you applications experience downtime of less than a minute.
  • Monitoring: Redis metrics are available through Monitoring, making it easy to monitor your Redis instances. Using Cloud Logging, you can also see the Redis logs for your instance.
  • On-demand billing: With Memorystore for Redis instances, you are billed by the hour for the capacity (GB) that you provision. If your instances are used for a few minutes, you are billed for only those minutes.
  • Redis versions 5.0, 4.0, and 3.2: The versions are always kept up to date with the latest critical patches, ensuring your Redis instances are secure. For more information on the versions and patch levels see the Memorystore for Redis Release Notes and Supported versions.

Connecting to a Memorystore for Redis instance

You can use any standard Redis client on the following environments to connect to your Memorystore for Redis instance:

  • App Engine flexible environment
  • App Engine standard environment
  • Compute Engine VM instances
  • Cloud Functions
  • Cloud Run
  • Google Kubernetes Engine clusters

Some serverless environments require a Serverless VPC Access connector as a prerequisite for connectivity with Memorystore for Redis. See Serverless VPC Access connector requirement for more details.

In order to connect to your instance, your client must be connected to the same network, and located in the same region as your Memorystore for Redis instance. See the Memorystore for Redis Networking page for more details.

Differences between managed and open source Redis

Memorystore for Redis is Redis protocol compliant. You can move your applications that use open source Redis to Google Cloud without any code changes.

In general, the Redis functionality that Memorystore for Redis instances provide is the same as the functionality that locally-hosted Redis instances provide. Some key differences are as follows:

  • Memorystore for Redis does not currently support Redis Clustering or Redis Sentinel.

  • Open source Redis provides multiple options for persisting data to disk. RDB persistence enables point in time snapshots of the dataset and AOF persistence logs every write operation received by the server. Memorystore for Redis currently does not support either RDB or AOF persistence.

  • Most parameters are preconfigured for Memorystore for Redis instances, and you cannot change them. Other parameters you configure when you set up your Memorystore for Redis instance. For more information, see Redis configurations.

  • Some commands can interfere with a managed Redis service and are blocked. For more information, see the list of blocked Redis commands.

  • Standard Tier instances do not allow reading from the replica.

Tier capabilities

The following table describes the differences between the available Memorystore for Redis service tiers:

Specifications Basic Tier Standard Tier
Description Provides a cache with no replication Provides redundancy and availability using replication
Max instance size 300 GB 300 GB
Max network bandwidth 12 Gbps 12 Gbps
Stackdriver Monitoring Yes Yes
Scale memory¹ Yes Yes
Cross-zone replication No Yes
Automatic Failover No Yes

¹ Basic Tier instances experience a downtime and a full cache flush during scaling up or down. Standard Tier instances experience minimal downtime and loss of some unreplicated data during the scaling operation. For more information, see Scaling Behavior.

What's next