What Grafana’s Purchase of Asserts.ai Means for the User
Grafana Labs‘ acquisition of Asserts.ai illustrates its aggressive expansion in what you receive with Grafana, especially with Grafana Cloud. Asserts’ purpose aligns with this expansion, offering easier and more autonomous analysis of metric data. Asserts simplifies this contextualization, a complex task for humans. This move demonstrates Grafana’s commitment to offering more to its paying customers, particularly as Grafana’s paid user base grows. Grafana competes with rivals like Datadog, Splunk and other observability platforms.
One distinguishing factor is Grafana’s non-retention of user data when used for non-commercial purposes. Crucially, Grafana’s expansion doesn’t compromise performance. It maintains its historical contributions and ensures compatibility with various observability open source tools, especially evident in its aggressive integration of Grafana with OpenTelemetry. Additionally, Grafana remains supportive of open source projects, including, notably, Prometheus, and continues to introduce new projects like the recent launch of Grafana Beyla, an eBPF auto-instrumentation tool.
In the immediate future, Grafana Cloud users will be able to benefit from the use of Asserts, which was created to help users find metrics data (or “contextualize” metrics data, as Grafana describes it) with the use of AI. It specifically scans the labels in Prometheus metrics and automatically discovers an application and infra components and how they’re connected to each other.
@grafana buys https://t.co/FXfBMAmFj7, to help simplify telemetry and observability. As @tom_wilkie describes Asserts, it « serves as a contextual layer for Prometheus metrics and provides insights into the relationships between system components. » @thenewstack #ObservabilityCon pic.twitter.com/DXy2bm9ams
— BC Gain (@bcamerongain) November 14, 2023
During Grafana’s ObservabilityCon 2023 annual users conference held in London this week, Asserts.ai founder and CEO Manoj Acharya — who was the fifth engineer AppDynamics hired — provided a look at what Asserts offers. He showed how the tool is used for scanning all the Prometheus metrics from a demo cluster he was running to build a service graph. “As it learns about new services, parts, and nodes, imagine you’re running a Kafka cluster; it will also acquire information on new topics,” Acharya said. Additionally, Asserts is being designed for Amazon integration, enabling it to read CloudWatch metrics and access dynamic tables, SQS queues, essentially forming a graph, which serves as an entity system for data, he said. “When thinking about understanding your data, consider a schema — an entity system now structured within a property graph. That’s what operates behind the scenes,” Acharya said. “It has currently discovered all the parts, nodes, and services in this specific cluster.”
Observability and Alerting
The significance of these capabilities becomes evident when considering the presence of around 20 diverse metrics solely measuring various memory consumption aspects, EMA analyst Torsten Volk said. Asserts.ai can automatically assess, using its dependency graph, which metrics are likely to impact a specific application and how frequently they need a collection to optimally feed the machine learning model. This results in quicker and more accurate predictions, potentially saving a significant amount of storage space. “Overall, Asserts.ai elevates Grafana from a data visualization platform to an application observability and alerting system, which is exciting news for both its current and prospective customers,” Volk said.
Indeed, Asserts.ai provides Grafana with the ability to offer “intelligent” dashboards that can automatically determine which metrics are actually relevant to predict, optimize and troubleshoot the performance and health of Kubernetes clusters. “This is no trivial task as Prometheus delivers a flood of metrics without distinguishing if and how they are relevant for a specific application. Ingesting, storing and analyzing all of these metrics can bury platform teams under an avalanche of data that needs to be ingested, analyzed and stored,” Volk said. “Asserts AI uses a combination of automated tracking of app and infrastructure dependencies, and machine learning to find the metrics that impact application health and performance and plots them on a Grafana dashboard. As Asserts.ai knows the interdependencies within the Kubernetes cluster, it can automatically detect upcoming problems and issue alerts that show the entire chain of issues all pointing toward the root cause.”
@grafana ´s Yasir Ekinci on the https://t.co/63mhrfZWlu purchase during his talk on AI/ML w/Ben Sully. Grafana Cloud users will be able to better determine not only problems with services, but « what are the components of your service? » at Grafana’s #ObservabilityCON in London pic.twitter.com/LVq6KQLaEl
— BC Gain (@bcamerongain) November 15, 2023
Grafana’s Yasir Ekinci, a senior software engineer, described during his talk at ObservabilityCon how Asserts would figure into the Grafana panel experience when asked during the Q&A session. His talk, with Grafana Senior Software Engineer Ben Sully, was on the role of AI and ML and how Grafana approaches generative AI. An example of that would be something like this: “What is going wrong with my service?” It’s not just about identifying your services; you also need to understand their components and dependencies. This might not be particularly relevant for generative AI but should be “more interesting” for Grafana Sift. This is because ”once you understand the service, especially its inbound and outbound directions, it becomes incredibly useful. Instead of just reaching your goals, you can get to the root of root causes faster,” Ekinci said.
Alex Murray, senior technology manager, reliability engineering, at UberEats competitor and Grafana customer Just Eat Takeaway, said Grafana’s Asserts purchase should serve to better interpret a deluge of metrics his team must often deal with. The company has operations worldwide and the challenge is to manage and monitor the magnitude of observability data in “a way that makes sense,” he said. “I see telemetry data as sort of a pseudo-identity in a data lake. So, you have a data lake of telemetry data, so you can cut and slice that data,” Murray said during the sidelines of the conference. “However, you want to give your business context, and to be able to get visualization and to learn things about your platform that you didn’t originally plan for.”
With Asserts, Just Eat’s operation team will be able to take advantage of another observability layer to integrate with the range of options Grafana’s panel offers, Murray said. “The really big thing with Asserts and Grafana is to turn your data into opportunity,” he said.
Additionally, the fact that Grafana does not require the organization to transfer data to Grafana’s server is a big plus, Murray said.
“It doesn’t matter where your data is. Unlike other observability providers, you don’t have to send everything to Grafana. With Grafana, it doesn’t really matter where your data is,” Murray said. He added that he also appreciated how Grafana offers ample compatibility with other data sources. “There are some other finer features like Correlations that allow you to link metrics together from different observability platforms.”