As Uplive grows, it has been periodically evaluating the offerings of the major cloud service providers. From a business perspective, Uplive places high priority on the stability of the underlying IT infrastructures, evaluating whether they can reliably support its long-term ambitions.
AWS has been the most important hosting platform for the Uplive app since its launch in 2016. According to Mingling Liu, CTO of Uplive, AWS was chosen for the following three reasons:
First, cutting-edge product and service offerings and wide infrastructure coverage allow Uplive to innovate quickly and easily to serve its growing global user base. AWS offers an extensive selection of customizable and user-friendly managed services, which seamlessly dock with and support the Uplive product team in terms of operating systems, programming languages, and web application platforms. This has freed the team from managing the complexities of the underlying system architectures to focus on business innovation. System maintenance has also become much easier compared with self-built solutions.
Second, for every milestone product update on Uplive’s end, AWS offers tailored product architecture and services that at once meet the new operational parameters and enable continuous cost optimizations. For example,
Amazon DynamoDB was employed soon after the launch of Uplive’s IM app to handle the high concurrent read/write requests, providing a minimum five-fold performance increase over the in-house MongoDB setup. Building on ,
Amazon Redshift, Uplive has tapped into the data analytics functions of
Amazon EMR,
Amazon Athena, and
Amazon Managed Streaming for Apache Kafka (MSK), to support operational decisions and user experience upgrades. Lastly, the company has also switched from
Amazon Relational Database Service (Amazon RDS) to
Amazon Aurora for its core database platform to take advantage of the latter’s powerful features.
“When choosing our technologies, we did a detailed comparison of almost all mainstream databases on the market. We found Aurora and DynamoDB to be a perfect fit for us, and were especially impressed by their price-to-performance ratio,” said Liu. “In particular, they allow our development team to remain essentially agnostic to the underlying system architecture, which is critical to achieving the fast-paced innovations demanded by web applications.”
Third, through AWS’ enterprise-class technical support, Uplive has access to technical experts and experienced engineers around the clock who can help Uplive to properly use the tools at its disposal and manage the IT environment itself. AWS also advises Uplive on its specific applications and use scenarios to optimize the system architecture. Furthermore, the AWS team has assigned Uplive a technical manager to coordinate resources and connect its team with technology specialists who can develop tailored optimization plans.
Uplive currently runs all the core services for its overseas live streaming platform on AWS, specifically Amazon Elastic Compute Cloud (Amazon EC2), Amazon Simple Storage Service (Amazon S3), Amazon ElastiCache, Amazon Virtual Private Cloud (Amazon VPC), Amazon DynamoDB, Amazon Aurora, Amazon Redshift, Amazon EMR, Amazon Athena, Amazon MSK, Amazon CloudFront, Amazon CloudWatch, Amazon Route 53, Amazon Global Accelerator, and AWS WAF.
Empowered by AWS, Uplive continues to deliver innovation and better experiences to its global users.
When Uplive first launched its products, it mainly relied on the basic AWS products—Amazon EC2, Amazon CloudFront, and AWS WAF—to enable secure, local access to services for users worldwide. Notably, DynamoDB helped Uplive to cope with traffic spikes with aplomb, as the company no longer needed to worry about how high the data volume could reach or devote development and maintenance resources to database partitioning and sharding as in traditional environments. In high concurrency read/write scenarios that create transient dataflows, such as for syncing offline messages, DynamoDB can still keep the latency to milliseconds.
Later, Uplive built a data lake architecture, together with autoscaling features such as that enabled by Spot instances, for further cost and efficiency improvements. Based on this architecture, Uplive has now created a complete data analytics and decision system that informs marketing and operational teams through near-real time data reports and full data visualization functionalities. This finer-grained management in turn enables Uplive to take the user experience “up a notch.”
As its global user base ballooned and new applications emerged, Uplive started to look for a core database service with better performance, lower running costs, and more user-friendly development support. The AWS team pointed the company to Aurora, which offers a wealth of powerful features such as compute/storage separation, multi-copy synchronization, instance read/write split, and high concurrency support. After an in-depth comparison with other relational databases including MySQL and PostgreSQL, the Uplive team found that Aurora could offer a minimum 3-5x performance gain under general workloads, and up to a ten-fold improvement in certain scenarios. For this reason, Uplive accelerated migrating its core business modules to Aurora, a decision that has greatly improved the reliability of its data and applications.
Following this migration, Uplive further tapped into Aurora Global Database for its multi-region data replication and disaster recovery functions. With this quick and cost-effective product, Uplive has achieved high availability for its entire system and set up multiple data processing centers for maximum business continuity. It can quickly rebuild its systems even following extreme scenarios.
AWS has allowed Uplive to develop a complete operations framework, an insightful data analytics and operational decision support system, and continuously improving cost optimization approaches. Amazon CloudWatch enables product developers and maintenance personnel to see changes in cloud resources in advance, and thus optimize instance selections and match resources with anticipated workload as early as possible. Aurora’s Performance Insights gives the team the ability to monitor database performance, define optimization areas, and predict and allocate resources accordingly to efficiently respond to potential workload spikes and achieve operational goals.
The AWS team has been involved in every stage of Uplive’s architectural refactoring, providing tools, methodologies, and industry best practices throughout its product development cycles to help it do more with less and become a leading global video social entertainment platform.
Liu is confident about what lies ahead: “Overall, development and innovation with AWS has been a smooth ride for us; the cost of learning the new tools is also low. AWS team’s efficient and relevant presentations have made our team more eager learners of new technologies and applications. We will continue to explore AWS products—for example, greater use of containerized applications—to bring innovations to the livestreaming, social networking, and other market areas.”