I recently received an automated reminder that I was approaching a work anniversary, which took me somewhat by surprise. It’s hard to believe that it’s already been a year (to the day) that I joined MongoDB! So I thought I’d take a moment to reflect on my MongoDB journey so far, share some exciting product updates, and signal where we’re headed next.
Our customers
I joined MongoDB because it built a product developers love. The innovation of MongoDB’s document model empowered developers to simply build. No longer encumbered by having to formalize and denormalize their data schema before their application was even designed, MongoDB enabled developers to interact with data in an intuitive JSON format, and made it easy to evolve data structures as the life of their application evolved.
One of my first steps upon joining the company was to learn more about our customers. I was excited to learn that in addition to delighting developers, MongoDB had launched capabilities that enabled it to win mission-critical workloads from enterprise class customers—including 70% of the Fortune 100 and highly regulated global financial institutions, health care providers, and government agencies. I found it remarkable that customers could replicate data across AWS, Google Cloud, and Microsoft Azure in MongoDB Atlas (our fully-managed cloud database service) with just a few mouse clicks, and that some customers replicate data between the cloud and on premises using MongoDB Enterprise Advanced. This optionality struck me as powerful in the era of rapid advancements in AI, as it enables customers to easily bring their data to the best cloud provider for AI.
Soon after I joined MongoDB, the team was firming up the development roadmap for the next version of MongoDB, and they asked for my input on the plan. The team was debating whether to focus on features developers would love, or governance capabilities required by large enterprises. I knew that ideally we would please all of our customers, so we had to try to make this an “and” and not an “or.” While I was new to MongoDB, from my 17+ years at AWS I learned that all customers demand security, durability, availability, and performance (in that order) from any modern technology offering. If a product or service doesn’t have those four elements, customers won’t buy whatever you’re selling. So as a team, we agreed that our next release of MongoDB—MongoDB 8.0—had to raise the bar for all of our customers, delivering great security, durability, availability, and performance.
The plan
We had less than a year before our target launch, so we knew we had to get moving, fast. My team and I brought MongoDB’s product and engineering organizations together to align on the plan for our next release. We set goals around delivering significant improvements in security, durability, and availability. And we set a line in the sand—that we weren’t going to release MongoDB 8.0 unless it was the best-performing version of MongoDB yet.
Measuring the performance of a feature-rich database like MongoDB can be tricky, as customers run a wide range of workloads. So we decided to run a suite of benchmarks to simulate customer workloads. We also developed Andon cord-inspired automation that would automatically roll back any code contributions that regressed our performance metrics. Finally, a set of senior engineering leaders met regularly to review our progress and immediately escalated any blockers that could jeopardize our launch, so that we could quickly fix things.
From my experience, I knew that great teams really respond when they’re given clear goals, and when they’re empowered to innovate, so I was excited to see what they would come up with. I’m proud to say that our product and engineering teams rose to the challenge.
Announcing MongoDB 8.0
Today, I’m thrilled to announce the general availability of MongoDB 8.0—the most secure, durable, available, and performant version of MongoDB yet! The team came up with architectural optimizations in MongoDB 8.0 that have significantly reduced memory usage and query times, and have made batch processing more efficient than previous versions.
Specifically, MongoDB 8.0 features:
-
36% better read throughput
-
56% faster bulk writes
-
20% faster concurrent writes during data replication
-
200% faster on complex aggregations of times series data
In making these improvements, we're seeing benchmarks for typical web applications perform 32% better overall. Here’s a breakdown of how MongoDB 8.0 performs against some of our benchmarks:
Improved performance benefits all users of applications built atop MongoDB, and for MongoDB customers, it can mean reduced costs (due to an improved price/performance ratio).
In addition to significant performance gains, MongoDB 8.0 delivers a wide range of improvements, including (but not limited to):
-
Improving availability by delivering sharding enhancements to distribute data across shards up to 50 times faster and at up to 50% lower starting cost, with reduced need for additional configuration or setup.
-
Improving support for a wide range of search and AI applications at higher scale and lower cost, via the delivery of quantized vectors—compressed representations of full-fidelity vectors—that require up to 96% less memory and are faster to retrieve while preserving accuracy.
-
Enabling customers to encrypt data at rest, in transit, and in use by expanding MongoDB’s Queryable Encryption to also support range queries. Queryable Encryption is a groundbreaking, industry-first innovation developed by the MongoDB Cryptography Research Group that allows customers to encrypt sensitive application data, store it securely as fully randomized encrypted data in the MongoDB database, and run expressive queries on the encrypted data—with no cryptography expertise required.
You might wonder why we’re so confident that customers are going to love MongoDB 8.0. Well, we’ve been acting as our own customer, and have moved our own applications over to 8.0. This approach is generally called “dogfooding,” but we think that “eating our own pizza” sounds more appetizing. Our internal build system—which our software developers use daily—is built atop MongoDB, and when we upgraded to MongoDB 8.0 we saw query latencies drop by approximately 75%! This was a double win, as it improved the performance of our own tooling, and it set our performance chat room abuzz with excitement in anticipation of delighting external customers. While results may vary based on your particular workload, the point is that we just couldn’t wait to share MongoDB 8.0’s performance gains with customers.
Indeed, customers are also already seeing great results on MongoDB 8.0. For example, Felix Horvat, Chief Technology Officer at OCELL, a climate technology company in Germany, said: “With MongoDB 8.0, we have seen an incredible boost in performance, with some of our queries running twice as fast as before. This improvement not only enhances our data processing capabilities but also aligns perfectly with our commitment to resource efficiency. By optimizing our backend operations, we can be more effective in our climate initiatives while conserving resources—a true reflection of our dedication to sustainable solutions.”
I encourage you to check out MongoDB 8.0 yourself. It’s available today via MongoDB Atlas, as part of MongoDB Enterprise Advanced for on-premises and hybrid deployments, and as a free download from mongodb.com/try with MongoDB Community Edition. In addition, customers upgrading from previous versions of MongoDB to 8.0 can find helpful upgrade guides on mongodb.com.
What’s next?
We’re excited for you to try MongoDB 8.0 and to share your feedback, as customer feedback helps us guide our roadmap for future releases.
Going forward, please watch this space. Over the next few weeks, we’ll be publishing a series of engineering blog posts that dig into MongoDB’s investments in the technology behind MongoDB 8.0. We’re also planning posts about horizontal scaling in MongoDB 8.0, and one that will look closely at queryable encryption (QE), but let me know what you’d like to hear more about.
It’s been an exciting year at MongoDB—I can’t wait to see what the next one has in store!
–Jim