
Ben Sigelman
@el_bhs
I like writing threads! Archive ➙ bit.ly/3s5QRI7
Co-founded @LightstepHQ.
Co-created @OpenTelemetry & @OpenTracing.
Co-created Dapper and Monarch at Google.
ID: 1543828177
http://bensigelman.org 24-06-2013 18:09:00
1,1K Tweet
5,5K Followers
566 Following

This isn't something I've thought about as much, but this look at the ballooning costs of #observability, via conversation between Ben Sigelman and 𝙳𝚘𝚝𝚊𝚗 𝙷𝚘𝚛𝚘𝚟𝚒𝚝𝚜 ✈️ ContainerDays 🇩🇪, is a really good primer with some excellent food for thought. buff.ly/2ExgVpc buff.ly/3JYsNAx

I wanted to write this months ago and procrastination got the better of me 😅but better late than never!! #observability can be fun and flexible - read about my own experience with OpenTelemetry


This is so timely! We did a #GameDay last week to assess sudden load spikes (not gradual). Our stack on #AWS has DynamoDB, #nlb, #ec2, and Pelikan Cache. All those parts worked - but we saw pressure in the most unexpected place: Throttling on our Lightstep is now ServiceNow Cloud Observability metrics.

Not news: for cloud-native apps, log mgmt is essential, and yet all too often it’s siloed and ROI-negative. "Yes news": the brilliant minds at Era Software are fixing that, and I am *THRILLED* to share ServiceNow’s plan to bring them to the Lightstep is now ServiceNow Cloud Observability party! servicenow.com/blogs/2022/clo…

Here's a stand-alone copy of the Golang auto-scaling OpenTelemetry base-2 exponential histogram github.com/lightstep/go-e… mapping and data structure. 100% of Lightstep is now ServiceNow Cloud Observability Go services are using this in production.

Today we launched Lightstep UQL (Unified Query Language)! As #observability shifts left, UQL helps #DevOps teams ensure #Kubernetes and other cloud-native applications are natively observable. Read more from Ben Sigelman ⬇️ #KubeCon #observabilityascode lightstep.com/blog/announcin…


I'm a big fan of the people and the overall approach at Momento -- moving more heavy-duty infra problems into specialized, fit-for-purpose solutions that improve performance, reduce risk, *and* reduce cost. (Congrats, Daniela Miao and Khawaja Shams!)



New video! What about #observability for OSS? How should you use OpenTelemetry to instrument a shared library? Plus some best design practices for writing stable software and backwards compatibility. Agree? Disagree? Let me know what you think. :) youtu.be/EAoKqdxHEjA



SPEAKER ANNOUNCEMENT 📢 Excited to kick off our official speaker announcements for #SLOconf with this talk by Adriana Villela | @[email protected] and 👩🏽💻✨Ana Margarita Medina ✨ of Lightstep is now ServiceNow Cloud Observability 🙌 What learnings should you leverage from the incidents you’ve been through? 🤔 Register sloconf.com


I really enjoyed this conversation with Daniela Miao the other day... we covered a lot of ground, including the early days of observability for distributed systems at Google, the best levers for AI in observability, the elephants in the room for TCO+ROI, etc. Worth a listen!

We are hiring for ML roles at Continua (Continua AI, Inc.)! Join our small AI startup and help bring our vision of personal agents to the world. continua.ai/careers Work with our small team (continua.ai/team) and have an outsized level of impact and autonomy. 🧵1/2



We had a chat with Ben Sigelman, co-founder of OpenTelemetry, about leading the industry to a brighter future with open source. Prior to OTel, observability required custom agents. OTel is leading the way toward a common instrumentation standard.
