Jump to Content
Databases

Go faster and cheaper with Memorystore for Memcached, now GA

March 1, 2021
Gopal Ashok

Product Manager, Memorystore

Ron Pontafaro

Product Manager, Databases

Last year, we announced the beta release of Memorystore for Memcached, a fully managed service compatible with open-source Memcached protocol. Memorystore for Memcached has delivered speed and cost savings for our customers and we’re announcing the general availability of Memorystore for Memcached. 

With Memorystore for Memcached, you can leverage the common Memcached in-memory key value store without taking care of upgrading, maintaining, patching, and more. You can scale based on your needs. In a common use case like database caching, you can lower not only your query latency but also the costs of your backend services. Since launch, we’ve heard from customers about improvements in such use cases of up to 5X in performance and up to 50% in database cost reduction.

Quizlet is one such customer who benefited from using Memorystore for Memcached. By migrating their self-managed Memcached cluster to Memorystore, they were able to lower costs by 50%. 

At Quizlet, site reliability engineers (SREs) have a saying: if Memcached goes down, Quizlet goes down. The reliability, seamless scaling, integration with Cloud Monitoring and the service being fully managed has made the life of SREs much easier according to Mason Leung, SRE at Quizlet. To learn more about Quizlet’s experience with Memorystore, check out their blog How Memorystore cut costs in half for Quizlet’s Memcached workloads.

Highlights of Memorystore for Memcached GA

Since preview, we have been hard at work to ensure the service is highly reliable and available. One of the key areas of improvement is self-service software updates. Memorystore for Memcached undergoes occasional maintenance, which requires node restarts. The automatic update is done over a six-hour window across all the nodes. This means the entire cache will be flushed over a period of six hours. This is usually suitable for the majority of applications, though in some cases you may need more control over the rate of cache flush during software update.

Self-service software updates allow you to have complete control of how mandatory software updates are applied to your Memcached instance. Using the self-service update API, you can control the time and rate at which updates are applied across the nodes in your Memcached instance. For example, if you have a 10-node instance, you can choose to apply the software update one node at a time over 10 days. Check out the documentation to learn more about self-service software updates.

The GA release also provides an uptime SLA of 99.9%.  

Memorystore provides a rich set of monitoring metrics. Here’s a look at the monitoring dashboard for an instance.

https://storage.googleapis.com/gweb-cloudblog-publish/images/memcached_prod.max-1100x1100.jpg
Click to enlarge

Low latency and performance is critical for our ad-serving platform. By using Memorystore for Memcached as a cache for our delivery service, we were able to improve p50 performance by 5x and reduce overall data management cost by 50%. Being a fully managed service, Memorystore has significantly reduced our operational cost

Mateusz Jeziorski, Chief Architect, OpenX

Getting started with Memorystore for Memcached

To get started with Memorystore for Memcached, check out the quick start guide. Sign up for a $300 credit to try Memorystore and the rest of Google Cloud services. You can start with the smallest instance and when you’re ready, you can easily scale up to serve performance-intensive applications. Enjoy your exploration of Google Cloud and Memorystore for Memcached.

Posted in