THL Simplifies Architecture with MongoDB Atlas Search

Nick Bell
October 7, 2024
#Atlas Search

Tourism Holdings Limited (THL) originally became a MongoDB customer in 2019, using MongoDB Atlas to help manage a wide variety of telematics data.

I was very excited to welcome Charbel Abdo, Solutions Architect for THL at MongoDB .local Sydney in July 2024 to hear more about how the company has significantly expanded its use of MongoDB.

The largest RV rental company in the world, THL has branches in New Zealand (where it is headquartered), Australia, the US, Canada, the UK and Europe. Specializing in building, renting, and selling camper vans, THL has a number of well-known brands under its umbrella.

In recent years, THL has made a number of significant digital transformation and technology stack optimization efforts, moving from a ‘bolt-on’ approach that necessitated the use of a distributed search and analytics engine to an integrated search solution with MongoDB Atlas.

THL operates a complex ecosystem managed by their in-house platform, Motek, which handles booking, pricing, fleet management, and more—with MongoDB Atlas as the central database.

Its +7,000 RVs are fitted with telematics devices that send information—such as location, high-speed events, engine problems, and geofences or restricted areas (for example, during the Australian bushfires of 2020)—to vehicles’ onboard computers.

THL initially used a bolt-on approach for complex search functionalities by extending their deployment footprint to include a stand-alone instance of Elasticsearch.

This setup, while functional, introduced significant data synchronization and performance issues, as well as increased maintenance overhead. Elasticsearch struggled under heavy loads which led to critical failures and system instability, resulting in THL experiencing frequent outages and data inconsistencies.

After two years of coping with these challenges, THL resolved to migrate away from ElasticSearch. After doing due diligence, they identified the MongoDB developer data platform’s integrated Search capabilities as the optimum solution.

"A couple of months later, we had migrated everything," said Abdo. "Kudos to the MongoDB account team. They were exceptional."

The migration process turned out to be relatively straightforward. By iteratively replacing Elasticsearch with MongoDB Atlas Search, THL was able to simplify its architecture, reduce costs, and eliminate the synchronization issues that had plagued the system. The simplification also led to significant performance and reliability improvements.

Because it no longer needed the dedicated sync resources processing millions upon millions of records per day, THL was able to turn off its Elasticsearch cluster and to consolidate its resources.

“All data sync related issues were gone, eliminated. But also we got our Friday afternoons back, which is always a good thing!” added Abdo.

Abdo’s team can now also use existing monitoring tools rather than having to set up something completely separate from the standalone search engine they were using.

“Sometimes, changes are easier than you think,” said Abdo. “We spent two-and-a-half years with our faulty solutions just looking for ways to patch up all the problems that we were having. We tried everything except actually looking into how much it would actually take to migrate. We wasted so much time, so much effort, so much money. While if we had thought about this a couple of years ago, it would have been a breeze.”

“Over-engineering is bad, simple is better,” he noted.

To learn more about how MongoDB Atlas Search can help you build or deepen your search capabilities, visit our MongoDB Atlas Search page.

← Previous

Vector Quantization: Scale Search & Generative AI Applications

We are excited to announce a robust set of vector quantization capabilities in MongoDB Atlas Vector Search . These capabilities will reduce vector sizes while preserving performance, enabling developers to build powerful semantic search and generative AI applications with more scale—and at a lower cost. In addition, unlike relational or niche vector databases, MongoDB’s flexible document model—coupled with quantized vectors—allows for greater agility in testing and deploying different embedding models quickly and easily. Support for scalar quantized vector ingestion is now generally available, and will be followed by several new releases in the coming weeks. Read on to learn how vector quantization works and visit our documentation to get started! The challenges of large-scale vector applications While the use of vectors has opened up a range of new possibilities , such as content summarization and sentiment analysis, natural language chatbots, and image generation, unlocking insights within unstructured data can require storing and searching through billions of vectors—which can quickly become infeasible. Vectors are effectively arrays of floating-point numbers representing unstructured information in a way that computers can understand (ranging from a few hundred to billions of arrays), and as the number of vectors increases, so does the index size required to search over them. As a result, large-scale vector-based applications using full-fidelity vectors often have high processing costs and slow query times, hindering their scalability and performance. Vector quantization for cost-effectiveness, scalability, and performance Vector quantization, a technique that compresses vectors while preserving their semantic similarity, offers a solution to this challenge. Imagine converting a full-color image into grayscale to reduce storage space on a computer. This involves simplifying each pixel's color information by grouping similar colors into primary color channels or "quantization bins," and then representing each pixel with a single value from its bin. The binned values are then used to create a new grayscale image with smaller size but retaining most original details, as shown in Figure 1. Figure 1: Illustration of quantizing an RGB image into grayscale Vector quantization works similarly, by shrinking full-fidelity vectors into fewer bits to significantly reduce memory and storage costs without compromising the important details. Maintaining this balance is critical, as search and AI applications need to deliver relevant insights to be useful. Two effective quantization methods are scalar (converting a float point into an integer) and binary (converting a float point into a single bit of 0 or 1). Current and upcoming quantization capabilities will empower developers to maximize the potential of Atlas Vector Search. The most impactful benefit of vector quantization is increased scalability and cost savings through reduced computing resources and efficient processing of vectors. And when combined with Search Nodes —MongoDB’s dedicated infrastructure for independent scalability through workload isolation and memory-optimized infrastructure for semantic search and generative AI workloads— vector quantization can further reduce costs and improve performance, even at the highest volume and scale to unlock more use cases. "Cohere is excited to be one of the first partners to support quantized vector ingestion in MongoDB Atlas,” said Nils Reimers, VP of AI Search at Cohere. “Embedding models, such as Cohere Embed v3, help enterprises see more accurate search results based on their own data sources. We’re looking forward to providing our joint customers with accurate, cost-effective applications for their needs.” In our tests, compared to full-fidelity vectors, BSON-type vectors —MongoDB’s JSON-like binary serialization format for efficient document storage—reduced storage size by 66% (from 41 GB to 14 GB). And as shown in Figures 2 and 3, the tests illustrate significant memory reduction (73% to 96% less) and latency improvements using quantized vectors, where scalar quantization preserves recall performance and binary quantization’s recall performance is maintained with rescoring–a process of evaluating a small subset of the quantized outputs against full-fidelity vectors to improve the accuracy of the search results. Figure 2: Significant storage reduction + good recall and latency performance with quantization on different embedding models Figure 3: Remarkable improvement in recall performance for binary quantization when combining with rescoring In addition, thanks to the reduced cost advantage, vector quantization facilitates more advanced, multiple vector use cases that would have been too computationally-taxing or cost-prohibitive to implement. For example, vector quantization can help users: Easily A/B test different embedding models using multiple vectors produced from the same source field during prototyping. MongoDB’s document model —coupled with quantized vectors—allows for greater agility at lower costs. The flexible document schema lets developers quickly deploy and compare embedding models’ results without the need to rebuild the index or provision an entirely new data model or set of infrastructure. Further improve the relevance of search results or context for large language models (LLMs) by incorporating vectors from multiple sources of relevance, such as different source fields (product descriptions, product images, etc.) embedded within the same or different models. How to get started, and what’s next Now, with support for the ingestion of scalar quantized vectors, developers can import and work with quantized vectors from their embedding model providers of choice (such as Cohere, Nomic, Jina, Mixedbread, and others)—directly in Atlas Vector Search. Read the documentation and tutorial to get started. And in the coming weeks, additional vector quantization features will equip developers with a comprehensive toolset for building and optimizing applications with quantized vectors: Support for ingestion of binary quantized vectors will enable further reduction of storage space, allowing for greater cost savings and giving developers the flexibility to choose the type of quantized vectors that best fits their requirements. Automatic quantization and rescoring will provide native capabilities for scalar quantization as well as binary quantization with rescoring in Atlas Vector Search, making it easier for developers to take full advantage of vector quantization within the platform. With support for quantized vectors in MongoDB Atlas Vector Search, you can build scalable and high-performing semantic search and generative AI applications with flexibility and cost-effectiveness. Check out these resources to get started documentation and tutorial . Head over to our quick-start guide to get started with Atlas Vector Search today.

October 7, 2024

Next →

MongoDB Database Observability: Integrating with Monitoring Tools

This post is the final in a three-part series on leveraging database observability. Welcome back to our series on Leveraging Database Observability! Our previous post showcased a real-world use case highlighting how MongoDB Atlas’s observability tools effectively tackle database performance challenges. Whether you’re a developer, DBA, or DevOps engineer, our mission is to empower you to harness the full potential of your data through our observability suite . Integrating Atlas metrics with your central enterprise observability tools can simplify your operations. By seamlessly working with popular observability tools, our approach helps teams streamline workflows and enhance visibility across systems. Integrating MongoDB Atlas with third-party monitoring tools MongoDB’s developer data platform combines all essential data services for building modern applications within a unified experience. Our purpose-built observability tools for Atlas environments offer automatic monitoring and optimization, guiding diagnostics tailored specifically for MongoDB. Additionally, we extend Atlas metrics into your existing enterprise observability stack, enabling seamless integration without replacing your current tools. This creates a consolidated, single-pane view that unifies Atlas telemetry with other tech and application metrics, ensuring comprehensive visibility into both database and full-stack performance. This integration empowers you to monitor, receive alerts, and make data-driven decisions within your existing workflows, driving greater efficiency. Below is a quick guide to modifying integration settings through the Atlas UI and the popular integrations we support: Navigate to the Project Integrations page in Atlas. Choose the organization and project you want to configure from the navigation bar. On the Project Integrations page, select the third-party services you’d like to integrate. Configure the chosen services with the required API keys and regions. Critical integrations for your observability platform With Atlas’s Datadog and Prometheus integrations, you can send critical MongoDB metrics to these platforms, empowering detailed, real-time monitoring. Through Datadog , you can track database operation counts, query efficiency, and resource usage, ideal for pinpointing bottlenecks and managing resources. Similarly, Prometheus enables you to monitor essential metrics like query times, connection rates, and memory usage, supporting flexible tracking of database health and performance. Both integrations facilitate proactive detection of issues, alert configuration for resource thresholds, and a cohesive view of Atlas data when visualized in Grafana. Atlas’s integration with PagerDuty streamlines incident management by sending metrics like performance alerts, billing anomalies, and security events directly to PagerDuty. This integration records incidents automatically, notifies teams upon alerts, and supports two-way syncing, ensuring resolved alerts in Atlas are reflected in PagerDuty. It enables efficient incident response and resource allocation to maintain system stability. With Atlas integrations for Microsoft Teams and Slack, you can route key metrics—such as query latency, disk usage, and throughput—to these channels for timely updates. Teams can use these insights for real-time performance monitoring, incident response, and collaboration. Notifications through these platforms ensure your team stays informed on database performance, storage health, and user activity changes as they occur. Use case: Centralized observability with MongoDB Atlas, Datadog, and Slack Let’s walk through a hypothetical scenario for ShopSmart, an e-commerce company that leverages MongoDB Atlas to manage its product catalog and customer data. As traffic surges, the DevOps team faces challenges in monitoring application performance and database health effectively. To tackle these challenges, the team leverages MongoDB Atlas’ integration with Datadog and Slack, creating a powerful observability ecosystem. Integrating MongoDB Atlas with Datadog: The team pushes key MongoDB Atlas metrics into Datadog, such as query performance, connection counts, and Atlas Vector Search metrics. With Datadog, they can visualize these metrics and correlate overall MongoDB performance with their other applications. Out-of-the-box monitors and dedicated dashboards allow the team to track metrics like throughput, average read/write latency, and current connections. This visibility helps pinpoint bottlenecks in real time, ensuring optimal database performance and improving overall application responsiveness. Setting up alerts in Datadog: The team configures alerts for critical metrics like high query latency and increased error rates. When thresholds are breached, Datadog instantly notifies the team. This proactive approach allows the team to address potential performance issues before they impact customers. Integrating Datadog with Slack: To ensure fast communication, alerts are sent directly to the dedicated Slack channel, “ShopSmart-Alerts.” This integration fosters seamless collaboration, enabling the team to discuss and resolve issues in real-time. With these integrations, ShopSmart’s engineering team can monitor performance quickly and address issues efficiently. The unified observability approach enhances operational efficiency, improves the customer experience, and supports ShopSmart’s competitive edge in the e-commerce industry. By leveraging MongoDB Atlas, Datadog, and Slack, the team ensures scalable performance and drives continuous innovation. Conclusion MongoDB Atlas empowers developers and organizations to achieve unparalleled observability and control over their database environments. By seamlessly integrating with central enterprise observability tools, Atlas enhances your ability to monitor performance metrics and ensures you can do so within your existing workflows. This means you can focus on building modern applications confidently, knowing you have the insights and alerts necessary to maintain optimal performance. Embrace the power of MongoDB Atlas and transform your approach to database management—because your applications can thrive when your data is observable. And that wraps up our Leveraging Database Observability series! We hope you learned something new and found value in these discussions. Sign up for MongoDB Atlas , our cloud database service, to see database observability in action. To dive deeper and expand your knowledge, check out this learning byte for more insights on the MongoDB observability suite and how it can enhance your database performance.

November 14, 2024