Chronosphere Flex Their Muscles

Welcome to edition #9 of the newsletter!

The big news this week is that full stack observability provider Chronosphere have snapped up the Calyptia log management platform. We also feature two products new to our radar - Quickwit a product that really lives up to its name, and TraceTest - a revolutionary way of harnessing OpenTelemetry traces. We also have exciting news on the Perses project, some major event announcements and some rather shady dealings with DataDog accounts.

Feedback

We love to hear your feedback. Let us know how we are doing at:

 [email protected]

https://twitter.com/TheObsGuy

NEWS

Chronosphere Snap Up Calyptia

We listed Chronosphere as ‘one to watch’ for 2024 and they have kicked off the year with a real statement of intent by acquiring Calyptia - a company built by the maintainers of FluentD and FluentBit. Calyptia Core is a sophisticated observability pipeline product with a large number of integrations. Strategically it is an astute move for providing an on-ramp from telemetry sources into the Chronosphere cloud space.

Perses Visualization Project Gathers Pace

The CNCF Perses project to create a standard for observability visualization is one that we have been following with interest. The latest release from the project includes a whole raft of updates - the most important being support for authentication with external OAuth/OIDC providers. The team have also been busy building an online demo and a new website where you can keep with up with all the latest Perses releases and documentation.

Monitoring as Code with Checkly

You are probably familiar with Infrastructure as Code and Pipelines as Code, now brace yourself for MaC - Monitoring as Code. The concept is being championed by synthetic monitoring specialists Checkly and forms the building blocks for their ‘programmable monitoring platform’. Rather than opting for a declarative approach with YAML/json, the framework is powered by TypeScript. This will involve a learning curve for those not familiar with Typescript, but the payoff is a much more natural and intuitive language for building conditional flows and looping constructs.

Sentry Announce Soft Rollout for Metrics

Sentry is well known for its error reporting and APM tooling. It has now announced that it will be adding metrics support to its product offering. The company’s stated aim is to build a metrics solution that is clearly focused on the needs of developers and they are inviting the community to participate in this GitHub discussion to help thrash out exactly what that means. There are no firm dates for release yet but there is a page on the company’s web site where you can sign up to receive updates on the progress of the rollout.

DataDog ‘resellers’ Get Collared

It is perhaps a sign of the times that a new racket doing the rounds is based on the lure of DataDog accounts pre-filled with six figure “credit” allowances. Last week, this reddit user posted an offer to “lease” their account for one year:

The post has now been removed by reddit admins. Meanwhile, a seller on Fiverr was offering an account with “100k credits” - a snip at just £2,890:

These ‘offers’ seem to play on the fact that DataDog does actually run a scheme offering credits to StartUps. Needless to say though, these ads are not legit and anybody taking a punt would end up getting a nasty bite.

Products

Quickwit - Log Searching at Warp Speed

The January meeting of the London Cloud Native MeetUp group included a presentation of Quickwit - a high performance log management and analytics tool - by company co-founder François Massot. QuickWit is a product that truly lives up to its name and is built from the ground up for blistering performance. The presentation showed a 4 vCPU server setup capable of ingesting 2TB of logs per day. The system’s query engine is also capable of searching through one billion logs in less than one second. As well as querying via the Quickwit CLI, you can also view analytics in the Quickwit Grafana plug-in.

Tracetest - Reinventing Testing With oTel

If you have implemented OpenTelemetry tracing on your distributed services, then Tracetest have a product which will enable you to leverage your investment and achieve massive time savings for QA's and devs in your organisations. TraceTest is a testing framework that builds out end-to-end testing processes just by hooking into your oTel traces. It effectively manages service discovery and actually allows devs and QA's to create asserts against spans in your OpenTelemetry trace store. There are already integrations with Grafana K6, Cypress and Vercel, with Playwright support imminent. They support a growing list of backends including Datadog, Honeycomb, Tempo, Sumo Logic, Jaeger and SigNoz.

From the Blogosphere

OpenTelemetry Tracing with Grafana Tempo

OpenTelemetry is redefining observability, but creating OpenTelemetry pipelines can be challenging. This comprehensive walkthrough from the Observability 360 web site provides step-by-step guidance in implementing an end-to-end pipeline for OpenTelemetry tracing on K8S. The article covers instrumenting .NET services to forward trace telemetry to the OpenTelemetry Collector. It also looks at configuring and deploying Grafana Tempo for trace storage and Grafana for querying and visualisation.

Monitoring K8S with the EFK Stack

Logstash vs FluentD/FluentBit is one of those great debates, like Ketchup vs mayonnaise on fries. This article from the RapidCode Technologies blog takes the FluentX path and includes a handy comparison of FluentD and FluentBit and when to use each of them. The authors go into into considerable detail on setting up and configuring the EFK stack (Elastic, FluentBit, Kibana) to provision an end-to-end K8S logging pipeline as well as covering important concerns such as filtering and indexing.

VIDEOS

Railway Build an Express Service with ClickHouse

Railway is a software deployment platform used by over 600k developers. In this brief video, one of the company’s engineers, Angelo Saraceno discusses how the company adopted ClickHouse as a backend store for the vast volumes of logs generated by the system. Their current architecture enables them to ingest 1.5bn logs per day whilst providing their users with an observability UI achieving query times of 50ms.

A Deep Dive With the Founders of VictoriaMetrics

In this lengthy but worthwhile video on the Geek Narrator YouTube channel, Kaivalya Apte talks to Aliaksandr Valialkin and Roman Khavronenko, co-founders of VictoriaMetrics. It discusses the story of how the company was set up, the architecture of the product and some of the technical intricacies of time series processing and storage. Many of their observations are of general relevance to managing time series data at scale.

Events/MeetUps

Dates have now been announced for two major events in Europe this spring. The CNCF Europe 2024 conference will be taking place in Paris from 19th-22nd March. This event includes the first ever ThanosCon. The following month sees Amsterdam playing host to GrafanaCon 2024 on April 9-10.

Dynatrace Perform, the annual get-together for the Dynatrace community, will be taking place in-person (in Las Vegas) and online from Jan 29th to Feb 1st.

If you are based in London, you will be pleased to hear that there is now an Observability Engineering MeetUp. It is also worth keeping an eye on the CloudNative London MeetUp. More globally you can get involved in the OpenTelemetry end user groups for Europe, Asia Pacific and the USA.

📣 Reminder!

Don’t forget - you can find a fuller listing of events on the Observability 360 calendar.

Quick Bytes

Karpenter is becoming an increasingly popular tool for Kubernetes resource management in the AWS world. There is now an Azure/Karpenter GitHub repo if you would like to try deploying it on an Azure AKS cluster.

If you are looking to skill up on eBPF, the CNCF have now launched a Cilium Certified Associate (CCA) certification. To help you on your way, the Isovalent team have put together this study guide.

OpenTelemetry can seem like a huge and daunting topic. This pending publication by Austin Parker, an oTel maintainer, and Ted Young, an oTel co-founder, promises to give you a solid grounding in the subject.

That’s all for this edition!

If you have friends or colleagues who may be interested in subscribing to the newsletter, then please share this link!

This week, our quote is from the great management theorist Peter Drucker