Blog

Testing distributed systems at scale is typically a costly yet necessary process. At Alluxio we take testing very seriously as organizations across the world rely on our technology, therefore, a problem we want to solve is how to test at scale without breaking the bank. In this blog we are going to show how the maintainers of the Alluxio open source project build and test our system at scale cost-effectively using public cloud infrastructure. We test with the most popular frameworks, such as Spark and Hive, and pervasive storage systems, such as HDFS and S3. Using Amazon AWS EC2, we are able to test 1000+ worker clusters, at a cost of about $16 per hour.

Netease Games is the operator for many popular online games in China like "World of Warcraft" and "Hearthstone". Netease Games also has developed quite a few popular games on its own such as "Fantasy Westward Journey 2", "Westward Journey 2", "World 3", "League of Immortals". The strong growth of the business drives the demand to build and maintain a data platform handling a massive amount of data and delivering insights promptly from the data. Given our data scale, it is very challenging to support high-performance ad-hoc queries to the data with results generated in a timely manner.

The Apache Spark + Alluxio stack is getting quite popular particularly for the unification of data access across S3 and HDFS. In addition, compute and storage are increasingly being separated causing larger latencies for queries. Alluxio is leveraged as compute-side virtual storage to improve performance. But to get the best performance, like any technology stack, you need to follow the best practices. This article provides the top 10 tips for performance tuning for real-world workloads when running Spark on Alluxio with data locality giving the most bang for the buck.

As the amount of data being collected and analyzed by Enterprises continues to grow unabated, more attention is being placed on managing the cost of storing the data relative to performance. Hadoop provides a scalable and fast way of storing and analyzing data, however, the cost of storing data in Hadoop is typically higher compared to alternative technologies like Object Stores.

From time to time, a question pops up on the user mailing list referencing job failures with the error message "java.lang.ClassNotFoundException: Class alluxio.hadoop.FileSystem not found". This post explains the reason for the failure and the solution to the issue when it occurs. This error indicates the Alluxio client is not available at runtime. This causes an exception when the job tries to access the Alluxio filesystem but fails to find the implementation of Alluxio client to connect to the service.

This blog describes our experience in speeding up Alluxio metadata operations using fingerprint and Alluxio under store bulk operations. These latest optimizations can be found in the 1.8.1 release. One of the major values Alluxio provides is a simple and unified interface to manage files and directories on different underlying storage systems. Alluxio acts as an intermediate layer and exposes a file interface for applications to interact with, even though the underlying storage system might be an object store that has a different interface.

we held our first New York City Alluxio Meetup! Work-Bench was very generous for hosting the Alluxio meetup in Manhattan. This was the first US Alluxio meetup outside of the Bay Area, so it was extremely exciting to get to meet Alluxio enthusiasts on the east coast! The meetup focused on users of Alluxio with different applications from Hive and Presto. As an introduction, Haoyuan Li (creator and founder of Alluxio) and Bin Fan (founding engineer of Alluxio) gave an overview of Alluxio and the new features and enhancements of the new v1.8.0 release.

This blog explores the challenges customers are facing with storing data long term in Hadoop, and discusses what the Hitachi Content Platform team is doing to help our customers solve these challenges with the help of Alluxio. Data is at the center of our digital world and for years Hadoop has been the go-to data processing platform because it is fast and scalable. While Hadoop has solved the data storage and processing problem for the last ~10 years, it achieves this by scaling storage and compute capacity in parallel. As a result, Hadoop environments have continued to expand compute capacity well beyond their needs as more and more of the storage is consumed by older, inactive data.

Recently, Qunar deployed Alluxio with Spark in production and found that Alluxio enables Spark streaming jobs to run 15x to 300x faster. In their case study, they described how Alluxio improved their system architecture, and mentioned that some existing Spark jobs would slow down or would never finish because they would run out of memory. After using Alluxio, those jobs were able to finish, because the data could be stored in Alluxio, instead of within Spark. In this blog, we show by saving RDDs in Alluxio, Alluxio can keep larger data sets in-memory for faster Spark applications, as well as enable sharing of RDDs across separate Spark applications.

Presto was designed from the ground up to offer interactive analytics using a massively parallel processing SQL engine that can combine data from multiple sources using a variety of connectors. As more and more companies discover the power of “separation of storage and compute” along with querying the data where it lies, it’s not wonder Presto is being asked to add even more functionality. Alluxio focuses its innovation at the data layer as a key enabling technology for Presto and a wide range of analytics applications and use cases. Performance is always critical, but providing memory speed response time is only part of the solution. If the application can’t access the data, it’s of no use.

Caching frequently used data in memory is not a new computing technique, however it is a concept that Alluxio has taken to the next level with the ability to aggregate data from multiple storage systems in a unified pool of memory. Alluxio capabilities extend further to intelligently managing the data within that virtual data layer. Tiered locality uses awareness of network topology and configurable policies to manage data placement for performance and cost optimizations. This feature is particularly useful with cloud deployments across multiple availability zones. It can also be useful for cost savings in environments where cross-zone or cross-location traffic is more expensive than intra-zone data traffic.