The push to cloud-native applications, and specifically to microservice landscapes, creates a challenge for operational teams. The challenge is, simply put, about sheer volume.
Humans cannot comprehend nor grasp the volume of operational telemetry coming out of metrics and monitoring systems. It simply doesn’t make sense to let humans interact with the raw telemetry.
Instead, monitoring systems tend to evaluate that data to create ‘normal’ states, only surfacing up anomalies from the nominal states. Chronosphere (chronosphere.io), a new player in the monitoring space, is no different. Chronosphere built their monitoring platform for complex and large-scale technology stacks. Their differentiation is the scale at which they can store and analyze telemetry. This gives customers more realistic and deep insights into their production environments.
“Being able to store tens of billions of metrics is only half the battle,” said Martin Mao, co-founder and CEO of Chronosphere. “While that’s a great first step that large enterprises across sectors will need, making sense of the data and using it effectively is even more challenging and that’s where our platform is focused.”
Chronosphere is founded by engineers that previously worked on large-scale systems at Uber and created the M3 monitoring platform. The deep experience that comes with monitoring such large and complex environment is the perfect motive to start your own company, it seems. Another notable example that went this route is Lightstep, where the founders had previously created Dapper at Google.
Tech
Chronosphere is built on the open source M3 as a SaaS-product, so companies can start using the monitoring platform within minutes and without any changes to their production environment. The first Chronosphere product is a hosted end-to-end monitoring platform for the enterprise, featuring security and access controls, resource management and rate limiting. It allows all teams within an organization to use a single, centralized platform. This cuts down on duplication of effort and removes much of the grunt-work traditionally associated with setting up a monitoring system yourself.
The Chronosphere UI showing a datapoint and thresholds
Chronosphere also provides visualization and alerting capabilities that are designed to provide for sane operability at scale given the volume of underlying data. The platform also helps auto-categorize the data and extracts higher level meaning for deeper insights.
The Chronosphere UI showing the embedded view for the linked Distributed Trace
The company’s technology is purpose-built for cloud-native microservice environments. The operational telemetry from the hundreds or thousands of services tends to be high cardinality or high dimensionality data. None of the open source tools, like statsd, Graphite, Carbon and ElasticSearch are able to keep up. At Uber, the team experienced limitations in all these products in terms of volume and complexity of monitoring data. This growth drove the movement from a few monolithic applications to thousands of microservices; as well as the migration from hundreds of servers to millions of containers.
Chronosphere is the resultant product of building a monitoring platform highly focused on scale at Uber, allowing efficient monitoring of microservices for large enterprises. Currently, the product is in private beta.
Competitors include Datadog, Lightstep and Sysdig.
Team
Martin Mao (CEO) and Rob Skillington (CTO) first worked together on the monitoring side of the initial launch of Office365. Later on, they built out the monitoring capabilities at Uber, eventually leading to the open source project M3 and founding Chronosphere.
Mao and Skillington led the team at Uber that scaled M3 to store tens of billions of time series, one of the largest production metrics systems in the world. The team clearly saw the benefits of engineering teams having such broad and deep insight into their applications and infrastructure and having the ability to make data-driven decisions quicker and more often.
They realized other high-growth organizations struggle with the same monitoring challenges when adopting a cloud-native architecture. They founded Chronosphere to bring the benefits of M3 to customers.
“At Chronosphere, the team is focused on providing something that companies can start using within minutes, while we continue to develop and maintain M3 as open source”, said Skillington.
Funding
In November of 2019, Chronosphere raised $11 million in funding. The round was led by Greylock Partners.
“Observability, and metrics in particular, have become the necessary pulse for all companies of all sizes. Metrics for apps, infrastructure, and business operations are now essential for companies to run their business”, said Jerry Chen of Greylock.
