April 2014 Summaries

Redis has become a popular backend host for job management systems such as Resque, Sidekiq, Celery, and Quartz, offering efficiency, low latency, and simple integration with PaaS providers like Heroku. The Redis JobStore implementation for Quartz provides a robust solution that can utilize almost any type of database as a backend, while also leveraging Redis's unique data types and commands to enable schema building and fast application requests. With the ability to instantly configure a database with Redis Cloud Platform, developers can simplify their job scheduling needs without the hassles of database setup and script writing. The Redis JobStore supports multiple schedulers and utilizes sorted sets and global locking mechanisms for quick and efficient trigger retrieval and execution. Overall, the integration of Redis with Quartz offers a powerful and scalable solution for large web applications that require quality performance and simplicity.

Today, Redis Labs announces a partnership with IBM's Cloud Marketplace, allowing customers to use their Redis Cloud and Memcached Cloud services immediately. This partnership enables the integration of these enterprise-class services with 100+ other SaaS applications and services in the marketplace, as well as SoftLayer Cloud. By leveraging IBM's expertise and track record, Redis Labs expands its reach while maintaining its core strengths of scalability, high availability, and stable performance. Customers can now easily access and try out these services through the IBM Cloud Marketplace, starting with a free 25MB plan that can be upgraded at any time without interruption.

This incident on Amazon's eu-west-1 data region highlights the importance of automatic failover mechanisms in Redis' services, which were able to keep customers with Multi-AZ plans unaffected and those using in-memory replication capabilities without downtime or data loss. The event also demonstrates the effectiveness of Redis' data persistence to EBS, allowing for quick data recovery, and its backup to S3 capability, enabling customers to quickly restore their data without intervention. The low number of customer contacts about this issue is encouraging, showcasing the protection offered by Redis' high availability mechanisms.

We are now offering Redis Cloud and Memcached Cloud services in two new data regions, ap-northeast-1 (Tokyo) and ap-southeast-2 (Sydney), which is a result of our expansion into the Asia-Pacific region. This expansion aims to provide users with improved performance by allowing their applications' database instances to be hosted closer to the application itself, reducing network latency and bandwidth limitations. Our cloud services offer infinite scalability, high availability, and stable top performance, with plans starting at 25MB for free, and can be upgraded instantly without interruption. The new data regions are now live, making it easier for users to access our enterprise-class Redis and Memcached solutions.

The process of comparing two versions of a database is used for testing, development, application updates, and new releases, providing advantages such as ensuring databases are fully synced, verifying backup restoration, and supporting master-slave replication. Redis uses this comparison to validate its replication mechanism between different versions and continuously validates synchronizations across cloud zones and regions for customers with managed databases. Initially, Redis RDB Tools was used, but it had drawbacks, including performance issues when dealing with large datasets. A script was written using Python's multiprocessing library to overcome these issues by parallelizing comparisons, printing discrepancies, and allowing improvements such as detailed output and expiration value comparison. This tool is suitable for constant data or real-time changes, but may not be ideal for altered data during the comparison process.

Today, SSL support has been added to Redsmin and a few patched Redis clients. Since Redis doesn't natively support secure communication, the responsibility falls on admins and developers to set up secure proxies like stunnel. To provide an easy and secure Redis environment, Redsmin now supports connecting, monitoring, and managing SSL-protected databases. Additionally, popular Redis clients have been modified to support SSL, reducing complexity in app operations. These modified clients are not yet merged into their respective trunks, but users can download the forks and provide feedback on their pull requests.

The company has taken steps to address the newly-disclosed Heartbleed vulnerability by upgrading their OpenSSL libraries and replacing certificates, which may have exposed user data due to a bug in the OpenSSL library. To prevent unauthorized access to their Redis accounts, users are advised to change all passwords and certificates, following specific steps for each service provider, including Heroku, CloudFoundry, AppHarbor, and AppFog. The changes require updating config vars or configuration variables with new credentials, which may cause applications to restart. Users should take these precautions to protect their accounts and data from potential exploitation of the Heartbleed vulnerability.

The known NoSQL vendor has announced the addition of in-memory capabilities to its database offering, sparking concerns about the hybrid approach. Redis, an open-source key-value store with solely in-memory capabilities, has shown that databases like it can significantly improve modern application performance by excelling at latency, throughput, and efficiency. Modern applications require sub-millisecond latency, instant scalability, and efficient data transfer, which in-memory databases like Redis can provide. Attempting to add in-memory capabilities to an incompatible database is doomed to fail, and a best-of-breed approach should be adopted where each component leverages the appropriate database for optimal performance.

The Redis team showcased its usage in the 2014 Game Developers Conference, highlighting its features such as sorted sets for leaderboards and session management for online applications, which are crucial for games that require real-time updates and responsiveness. The efficient hash data structure makes it easy to manage player data, profiles, and other game-related information. Many developers reported using Redis in conjunction with other datastores like MySQL, Cassandra, and MongoDB to augment its capabilities and handle large amounts of numerical data.

Last week, Google's Urs Hölzle announced a significant price reduction for Google Cloud Platform services during an event in San Francisco. The following day, Amazon's Andy Jassy revealed the company's 42nd price cut at AWS Summit keynote, reducing costs for various cloud services by 10-40%. In response to these announcements, Redis Labs has also reduced prices of all plans on AWS by 10-40%, with the new pricing taking effect immediately. The company's fully-managed and all-inclusive service model includes factors beyond RAM costs, such as EBS storage, I/O, and network expenses.