AWS Partner Network (APN) Blog
Aerospike Cloud: Scalable Real-time Database-as-a-Service with AWS SaaS Factory Support
By Lenley Hensarling, Chief Product Officer – Aerospike
By Akshay Patel, Senior Practice Lead, SaaS Partners – AWS SaaS Factory
By Anthony McClure, Senior Partner Solutions Architect – AWS SaaS Factory
By Bill Tarr, Principal Solutions Architect – AWS SaaS Factory
 |
Aerospike |
Real-time, high-performance workloads demand low-latency databases that can scale to meet the most stringent requirements. Distributed databases capable of delivering on these demands have been available for some time. However, these databases can be operationally burdensome to manage. As these workloads transition to the cloud, the complexity of managing the required infrastructure can increase, posing challenges for customers.
Aerospike has been managing databases in the cloud for years and has leveraged key insights from those operational experiences to build its next-generation database service, Aerospike Cloud. By delivering Aerospike as a software-as-a-service (SaaS), the company has eliminated the need for customers to allocate infrastructure, deploy and manage new software, and train staff on operational aspects. This SaaS model enables faster adoption of Aerospike’s capabilities by customers.
The AWS SaaS Factory team invited Lenley Hensarling, Aerospike’s Chief Product Officer, to discuss the Aerospike’s journey to SaaS. Lenley joined Aerospike in 2019 as Chief Strategy Officer and has been leading the Products organization for the past three years. Before that, he had extensive experience across engineering management, product management, and operational management at both startups and large successful software companies. He has worked with enterprise applications and infrastructure software, with a focus on real-time data and applications.
Q&A with Aerospike
AWS SaaS Factory: Tell us about Aerospike’s products and solutions overall.
Lenley Hensarling: Aerospike has been a leader in delivering database performance at scale at an efficient cost for over a decade. Our distributed, multi-model, NoSQL platform provides unmatched speed (sub-millisecond latency) and reliability with strongly consistent transactions. In the last few years, our journey has expanded our product development to be cloud-first as we launched Aerospike Cloud. Our managed service set the groundwork for this.
Aerospike is deployed for a range of real-time use cases including payments, fraud detection, personalization, and risk management. Aerospike is often core to our customers’ AI/ML pipelines. Last year we launched our Aerospike Graph Service which enables customers to query very large graphs in real time. We recently have made available our Aerospike Vector Search product which supports use cases including Retrieval Augmented Generation (RAG), Semantic Search, and caching of results from Large Language Models (LLMs) and Small Language Models (SLMs). As enterprises make use of Generative AI, they are going to need a vector solution with the throughput required for deployment at scale, so we added Aerospike Vector Search, built on our foundational database.
AWS SaaS Factory: How does having a SaaS offering in Aerospike Cloud impact your other products?
Lenley Hensarling: Historically, many of our customers came to us when they hit the scaling tipping point when the performance and economics of their prior solution made it prohibitive to continue. Now, developers can more easily build their apps from day 1 with Aerospike and not worry about hitting that cost or performance wall.
These developers come from all verticals, with transactional systems that need consistent and reliable performance as their throughput extends to millions of transactions per second or data storage grows from gigabytes to hundreds of terabytes and on to petabytes.
Aerospike Cloud provides a simple access point to sign up, pay, deploy, and manage Aerospike databases. From there, a developer can connect to their application securely over https or establish a private link between their application and the database.
AWS SaaS Factory: Can you walk us through how the SaaS Factory team supported the adoption of a SaaS model?
Lenley Hensarling: The SaaS Factory team has been a great source of knowledge for us to use while designing and building our services. From architecture design reviews to discussions about pricing and costing models, their insights and advice accelerated our learning curve. We were fortunate to consider the SaaS Factory as an extension of our team.
AWS SaaS Factory: What process or organizational changes were required as you tried to deploy a database as a service?
Lenley Hensarling: We went through some changes, but we were starting from an advantageous position. Aerospike is used by Fortune 100 companies to power mission-critical applications. Companies like Airtel, Fidelity, PayPal, and Snap rely on the Aerospike database to win in their respective markets. Aerospike had an organizational structure to assist and support these customers in successfully delivering these applications. Adding a SaaS delivery model, we were able to extend what we were already doing to help these customers with their Tier 0 and Tier 1 applications.
One of the interesting challenges was managing different dependencies within Aerospike Cloud. Aerospike has many engineering groups for our servers, clients, and tools. Aerospike Cloud leverages several of these components in our data planes. With these cross-team dependencies, we needed a systematic dependency and risk management process to drive our change control process. This reinforced that the evolution was a company-wide collaborative effort, not just a single initiative.
AWS SaaS Factory: What were some of the technical challenges in the move to a SaaS delivery model?
Lenley Hensarling: Aerospike’s database has always excelled at speed at scale, yielding high throughput workloads with sub-millisecond latencies for reads and low single-digit latencies for writes. To successfully deliver SaaS, we needed our solution to be better cost optimized than alternative approaches. We deliver on this based on a few characteristics of our architecture, including effective use of modern hardware. AWS provides Aerospike with a great set of Amazon EC2 instances to enable enterprise deployments.
Aerospike and other NoSQL databases provide high availability and resiliency through their distributed architectures. In the cloud, this means having a single database span across AWS Regions and availability zones. This is another example of how capabilities and cost structure vary in a SaaS environment.
We also developed a cost model for distinct parts of our architecture. This included digging into variances in data transfer costs of write operations across various replication factor values in a multi-zone database and understanding the data availability SLAs across various replication factor values, to name just a few.
AWS SaaS Factory: What feedback are you hearing from customers about addressing those challenges in the move to a SaaS delivery model?
Lenley Hensarling: The feedback has been great. Existing on-prem and cloud managed service customers have appreciated having the option to quickly spin up new database clusters on the SaaS platform. New users appreciate the ease of use of our free trial and the underlying price-performance capabilities. We have had some feature requests for things like expanded observability metrics and allowing customers to purchase directly from the AWS Marketplace. Most of these requests are for features that have been on our roadmap and it’s always good to have validation from our users that we are on the right track.
One important shift we saw in this transition is the mix of personas we engage with. The shift to SaaS has changed the typical user from the operations to the developer role. This changed the type of input and questions we hear. Developers are focused far more on their apps than the operation of the database. That’s a good thing! Aerospike Cloud was designed so that our developers don’t need to worry about any of these operational procedures. This has changed how we support free trials and onboard new customers. We’ve invested heavily to provide the resources to address different developer use cases like integrations, developer-focused documentation, and code samples.
AWS SaaS Factory: Can you share how AWS SaaS Factory helped you address these technical challenges?
Lenley Hensarling: AWS SaaS Factory was a true partner in this journey. Having access to their experience with many AWS partners has been a great help. We had regular milestone reviews with their team to dive into the details of each component in our architecture and discuss the pros and cons of various approaches. In some cases, we had come to a build versus buy decision and SaaS Factory provided recommendations for different 3rd party tools and libraries to include in our evaluation.
We also appreciated their approach to working with our team. AWS has an extensive catalog of services that you can leverage when building software. AWS SaaS Factory took extra time to understand the problem we were trying to solve and when they arrived at a recommendation, we received an objective point of view that was easy for the team to understand.
AWS SaaS Factory: Well, thank you for the partnership, and congratulations on your recent successes. Any final advice for other software providers as they navigate their SaaS journey?
Lenley Hensarling: Do as much planning ahead as you can. Defining your goals, your target audiences, and your internal stakeholders early is essential as you face challenges and decisions on the way. This helped us prioritize where we sought user feedback. It crystallized our timeline for delivering a minimum viable product (MVP) in a compressed amount of time. It helped prioritize our resources and ensure peak database performance and cost efficiency. While we’re excited about this success, this new platform serves as the focal point for Aerospike’s cloud-first strategy. We can’t wait for you to experience some great new AI and ML capabilities in 2024 including Aerospike Vector Search.
About AWS SaaS Factory
AWS SaaS Factory helps organizations at any stage of the SaaS journey. Whether looking to build new products, migrate existing applications, or optimize SaaS solutions on AWS, we can help. Visit the SaaS on AWS page to discover more technical and business content and best practices.
SaaS builders are encouraged to reach out to their account representative to inquire about engagement models and to work with the AWS SaaS Factory team.
.
Aerospike – AWS Partner Spotlight
The multi-model, NoSQL real-time data platform for multi-cloud, large-scale JSON and SQL use cases. The Aerospike database has the lowest latency with predictable performance at any scale, with industry-leading uptime, requiring up to 80% less infrastructure.