Hey everyone <!here>, we’re going to be at the Zero Downtime Hackathon in San Francisco this Saturday, August 22_._ 🚀 The theme is building reliable AI systems, with SigNoz powering observability. You’ll spend the day building a working app, pipeline, or agentic workflow, and use SigNoz to monitor traces, metrics, logs, latency, and failures as part of the project. If you’re in SF, come by, build something cool, and say hi to the SigNoz team Register here: luma.com/zero-downtime

Hey folks 👋 We’ve made some of the biggest changes to SigNoz in the past few months. All around one question: what should observability become in the AI era? We believe the observability platform of the AI era won’t be today’s platform with an MCP server attached. Giving agents access to logs, traces, and metrics is just the starting point. The harder problems are deeper: → Can agents reason over telemetry without burning too many tokens? → With an explosion of telemetry from agentic workloads, can platforms keep up without slowing down? → How do humans stay in control as traces and workflows get more complex? These questions have shaped our roadmap. In the last few months, we’ve come a long way building toward this vision. Full writeup on where we think observability is headed and what we’ve been building: signoz.io/blog/observability-for-the-ai-era Curious what changes you’re seeing in this space 👀

hey <!channel> I'm back with another interface update I want you folks to take a look at and shape. The first image is where we have 2 separate mechanics on the tooltip - Click to see drilldown options and press

P

to pin. The second demo video is the approach I have been considering, where just clicking on the relevant point does both for you together. Happy to hear detailed feedback, but if not, press 1️⃣ for a combined mechanic, and 2️⃣ to retain the existing one

Hello folks <!here> ! The Agents of SigNoz Hackathon has officially BEGUN! This week, developers around the world are building agent-native observability with SigNoz. Instrument your applications with OpenTelemetry, collect traces, metrics, logs, and build dashboards that help you understand exactly what's happening inside your AI agents and applications. Build SRE copilots, AI/LLM observability workflows, or anything that helps teams see deeper into production. Prizes 👇🏻 $20,000+ in prizes Job interviews at @SignozHQ You can participate solo or in teams of up to 4, and every member of a winning team receives the full prize. Not registered yet? There's still time to join: forms.gle/… Join #C0BE9PHMHD2 to receive regular updates and reminders for the hackathon.

Hey folks <!here>! 👋 Excited to announce that we are hosting a hackathon with $20,000 in prizes! Build agent-native observability workflows, SRE copilots and AI/LLM dashboards using SigNoz and win cool prizes 🥳 Find more details below 👇🏻 Register here: forms.gle/… Prizes, credits and benefits worth $20,000: - 💻 Apple MacBooks, iPads and iPhones to every team member - 💰 $10,000 in AWS credits - 💸 $100 free AWS credits for every participant - 🧑‍💻 Job interviews at SigNoz for top winners - 🏎️‍ LEGO Ferrari SF-24 set worth $250 for best blogs - 🎁 Exclusive swag for top social media posts 🌏 Complete details on website: wemakedevs.org/hackathons/signoz Join #C0BE9PHMHD2 to receive regular updates and reminders for the hackathon.

Hey everyone 👋 If you are starting with OpenTelemetry, just dropped a video that covers everything from the ground up. Here is what we go through: → How OpenTelemetry came to be and what problems it was actually solving → The architecture and how everything connects together → A proper deep dive on the Collector → Clearing up the API vs SDK confusion → A live demo using the opentelemetry-demo-lite project — traces, metrics and logs flowing into a SigNoz dashboard in real time Everything straight from the official docs and covered within 15 minutes. Good one to share with anyone on your team who is just getting started with OTel. Would love any feedback from the community. More videos in the series coming soon, with next up being AI observability 🙌🏻

Sharing a writeup that digs into something a lot of us are starting to think about: what happens when AI agents become a second user of our observability stack, alongside humans? 👀 The piece follows Leo Blondel, CTO at a 3-engineer startup, who plugged Claude Opus into his telemetry via MCP and let it run as an always-on triage agent. Architecture is interesting: • Alert rules stay deterministic and human-tuned. • Agent receives webhook → queries telemetry through MCP → cross-references k8s state + recent GitLab merges → only escalates via Slack if something looks real. • Postgres for working memory, daily restart to manage context windows, sub-agents with fresh contexts for the "dirty work" so the main agent stays clean 3 weeks running, zero false positives, one false negative. The framing I found most useful: agents bring context, humans bring care. An agent can read every log line and correlate every trace, but it doesn't know which 500 actually matters for the business this week. Keep humans on judgment, let agents do the grunt work below. There's also a solid section on what this asks of a platform which unified data model, MCP-style interfaces, legibility to models trained on open docs/OTel. Worth reading even if you're not on SigNoz specifically. https://newsletter.signoz.io/p/your-observability-tool-just-got

Hi! This week's edition of Observability Real Talk answers a question most engineers file under "I'll deal with this later": why is a trace ID exactly 128 bits long? 👾 Here are the major plots covered, • Why trace IDs can't be incremental counters, coordination is too expensive when every service generates IDs independently • The birthday paradox, and why our intuition about collision probability fails so badly • The math behind

P(collision) ≈ 1−e^(-k²/2N)

, broken into three acts: Safe, Danger, Inevitable • Why 64 bits gets you to a 2.7% collision risk at 1 billion IDs, and nearly 100% by 10 billion • Why 128 bits keeps the collision cliff beyond any realistic horizon • Why 256 bits is mathematically better but practically wasteful I write deep, in-detail guides on OTel every week here, so make sure to subscribe if you wish to follow along!

Most developers and engineers who start experimenting with observability usually begin with VM-based setups. That’s also where things can start getting messy especially when setting up OpenTelemetry and ensuring everything works end-to-end. To make this easier, we’ve built SigNoz VM Collection Agents to simplify the setup on VMs. We’ve also put together a walkthrough covering: • how the setup works • configuring VM agents • running the OpenTelemetry collector • validating telemetry inside SigNoz If you’re getting started with VMs you have the green flag ! Watch here:

Hey everyone 👋 We’ve officially wrapped up the full collection agents series with the final environment being ECS (EC2), now covered in a 2-part release. These last two videos focus specifically on setting up observability on ECS using SigNoz collection agents, completing the entire series end-to-end. If you haven’t tried it yet, I’d definitely suggest spinning up a small setup on AWS, experimenting with ECS, and following along it should be pretty seamless to get running. Going forward, I’ll be working on a new set of videos. If there are any specific areas or use cases you’d like me to cover next, feel free to drop your suggestions here. Both ECS parts are now live would love to hear your thoughts ! Part 1: Part 2:

Hey folks! 👋 SigNoz is now officially on the Vercel Marketplace! 🎉 If you're deploying on Vercel, you can now connect your projects to SigNoz in one click — no manual drain setup, no custom headers. What you get automatically: → Logs from Lambda, Edge, Static, Build & Firewall sources → Distributed traces via Vercel's native OTel trace drain → Full APM — latency, error rates, p50/p90/p99 → LLM observability if you're using the Vercel AI SDK Since SigNoz is built natively on OpenTelemetry, Vercel's trace drain connects without any translation layer or proprietary agents. 🔗 Install: vercel.com/marketplace/signoz 📖 Docs: signoz.io/docs/integrations/vercel/ Would love to hear from anyone already using Vercel — happy to help you get set up! 🙌

Hi! This week's edition of Observability Real Talk covers how our platform team rebuilt our internal observability after the original system crashed under its own telemetry volume. ❤️‍🔥 Here are the major plots covered, • Why our V1 setup (Testbed + SaaSMonitor) kept choking and why adding Envoy as a load balancer didn't save it • The three-role split: Builders (daemonsets per node), Rangers (cluster + endpoint probing), Stewards (the buffered pipeline) • Annotation-driven per-container scraping, built ~6 months before OTel considered it • Persisting Kubernetes events past the default ~1h retention • The Envoy → OTel Gateway → Redpanda pipeline that finally held • How we use nearly all 7 OTel Collector deployment patterns in one system • The Castle, our dedicated SigNoz instance watching 6 regions, 21B metric points/day I write deep, in-detail guides on OTel every week here, so make sure to subscribe if you wish to follow along!

Hey everyone, We’re actively expanding our YouTube with more hands-on videos around SigNoz Collection Agents - the latest one dives into Docker and how you can simplify your OpenTelemetry setup there. Alongside this, we’re also covering observability across different ecosystems - whether it’s tools like Agno, newer trends like OpenClaw, or setups around AI workflows and Hugging Face. The idea is to explore how observability fits into modern systems end to end. If you’re building or experimenting with your own setups, these videos should help you get a clearer picture and improve how you monitor things in practice. Do check out the channel - more interesting videos coming soon. Sharing the latest one here 👇 https://www.youtube.com/@signoz/videos

Hey everyone!! 👋 We’ve been seeing a common pattern where OpenTelemetry configs start getting quite complex once you move beyond basic setups. A lot of wiring, different components, and it’s not always easy to piece everything together. We’re working on a video playlist around this using SigNoz Collection Agents, mainly to help simplify how these configurations are structured and understood. We’ve just published an introductory video where I walk through a simple e-commerce example to show how observability grows in a real setup and what kind of telemetry + wiring you usually have to deal with. The idea is to help connect the problems you might be facing with a more structured way to approach them. Here’s the intro video: We’ll be covering more setups next, including Docker, Kubernetes, and others, going step by step. Would love to hear your feedback or any specific issues you’ve faced. Happy to shape upcoming videos based on that 🙌🏻

Hi! This week's edition of Observability Real Talk about how companies like Uber and Netflix observe their massive systems all day. 👀 Some of the major points covered include, • How Netflix shifted from Epic to Atlas • Rules of thumb followed by Netflix for storing telemetry data • How Netflix built Atlas Streaming Eval • How Uber shifted from graphite stack to M3DB for metrics • How they built Jaeger for distributed tracing etc Next drop would be on how we at SigNoz think and built our observability systems for scale, subscribe to stay tuned!

Hi @Jeethu Karthik - Why is that your have assistant userexperiance in signoz UI? Many users lives in team channels/slacks channel. Why do you want to ask your users to log in to UI for this?

We deploy SigNoz to Kubernetes using Helm. In that case, what advantages would we gain from using Foundry?

Hi! This week's edition of the Observability Real Talk covered how we built a high-performance JSON log store, overcoming Clickhouse's JSON constraints. We are touching up on, - Some of the problems we faced - Exploring max_dynamic_path option setting - How we built a 2-tier log storage system, which drastically improved our efficiency Lmk your thoughts and subscribe if you love such deep engineering lore! https://newsletter.signoz.io/p/overcoming-clickhouses-json-constraints

Hi! In this week's edition of Observability Real Talk newsletter I covered the three main deployment patterns for OTel Collector at Scale. - Load balancer pattern - Multi-cluster pattern - Per-signal pattern I've also added tips on choosing your deployment pattern based on your architecture, as well as some first-hand advice from an OpenTelemetry contributor! Let me know if you enjoyed this also do subscribe! 😉 Link - https://newsletter.signoz.io/p/patterns-for-deploying-otel-collector

Hey guys! 🤓 I'm Elizabeth (DevRel at SigNoz) and write for the observability real talk newsletter and this week's edition we covered about Auth for Otel Collector - something crucial but often overlooked. Collectors are THE most busy data hubs of your observability pipeline, handling huge amounts of information every moment, hence securing them is non-negotiable. Here are some main takeaways, 1. Main points of data movement in the collector 2. Learning more about extensions 3. Securing incoming traffic with bearer auth etc 4. Securing outgoing traffic with sigv4auth Give it a read and hit the subscribe button! https://newsletter.signoz.io/p/eli5-auth-model-for-opentelemetry

Hello Folks 👋 <!here> We have just shipped our 100th open source release! This journey would not have been possible without your amazing support, insightful feedback, and valuable contributions. You are truly at the heart of everything we build, and we're so grateful to have you with us. To celebrate this milestone, we’re thrilled to introduce you to our new friend and official mascot! Say hello to Olly! We hope you love the new artwork and stickers as much as we do. The celebrations don't stop here! We'll be at KubeCon + CloudNativeCon North America in Atlanta, Georgia, from November 10 to 13, 2025. If you're attending, please stop by booth #1372 to say hi, meet the team, and grab some cool merch. We'd love to see you there! From all of us at SigNoz, thank you for being on this incredible journey. We're just getting started, and we can't wait to build the next 100 releases with you. With immense gratitude, Nagesh

Hey <!channel>, we're re-working our trace details page and wanted to give you a little sneak peek at what we're cooking and also open this for your feedback early in the process. Let us know if this approach serves your use case better, and what we can fix/improve. Looking forward to releasing this in the wild for y'all to use 🤘

hey <!here> 👋 We are looking to hire Forward Deployed Engineers based in the US. If you are an SRE/DevOps and like working with new problems & customers - this role may be a good fit for you. If you know anyone who may be interested in this role, please share this with them. 🙏 Here's the JD and application link - https://jobs.ashbyhq.com/SigNoz/8f0a2404-ae99-4e27-9127-3bd65843d36f?locationId=0b7e5201-9671-4ce7-b000-2941cfb016e9

Hey <!here>! 👋 Exciting news! SigNoz is now live on the Vultr Marketplace You can now deploy SigNoz with a single click and get full-stack observability up and running in minutes, no manual setup needed. Give it a try and explore how easy it is to monitor your applications with SigNoz. We’d love your feedback on the experience. Let us know what you think!

Hey <!here>! 👋 Quick reminder - Day 3 (Wednesday, Sept 10) of Launch Week 5 is all about our OSS community! You asked for production-ready database options beyond SQLite - PostgreSQL support is now live! But that's just the start. We've rebuilt our entire SQL layer to be database-agnostic, so MySQL/MariaDB are coming next based on your feedback. What else we're covering? • Live demo of the new PostgreSQL setup • Railway one-click deployment (yes, actually one click!) • How we redesigned our schema for scale • What's next for making data ingestion easier This entire day exists because of your feedback. Come tell us what you want to see next! Register if you haven't already: https://lu.ma/ockltj0t

Hello <!here>! 👋 We're running Launch Week 5 from September 8-12, dropping 5 major features - one each day. Special highlight for our OSS Community - Day 3 (Sept 10) is all about you: https://lu.ma/ockltj0t • Choose between PostgreSQL and SQLite for your deployment • One-click deployment through Digital Ocean Marketplace • Features specifically requested by the community The entire week's lineup: Day 1: Interactive Dashboards - drill down without opening new tabs Day 2: Query Builder v5 - SQL-like syntax with auto-completion Day 4: Trace Operators - query service relationships directly Day 5: Cost Control - track and manage observability costs Full schedule: https://lu.ma/signoz-launch-week Each session is 30 minutes, and features go live immediately after the demo. Join the ones that matter to you. See you there!

Hello folks!! Starting *August 28, 202*5, Bitnami will change its service model (see Bitnami/charts#35164), SigNoz depends on Bitnami container images and Helm charts for Zookeeper. We are taking measures to ensure continuous service for our community. Our progress so far: • Container Images: We have migrated Zookeeper Bitnami images to SigNoz-managed repositories: signoz/zookeeper. A copy of the current Zookeeper Bitnami images has been retained, although no new releases or builds will be supported. • Helm Charts: We have extracted and maintained the Bitnami Zookeeper chart within the SigNoz repository (PR #725). For you, this means: • Most users: No disruption; things will continue to work as expected. • If you use custom deployments with direct Bitnami chart references or manage your own SigNoz infrastructure, please update to the latest version before August 28, 2025. For more information, please refer to Migration from Bitnami Dependencies

Hello <!here>, We're going live in about 2 hours with a fireside chat. Pranay will be leading a conversation with our friend Austin Born (Founder of Shinzo) on Making LLMs Observable. If this topic intrigues you, please drop by and ask questions. We start at 9 AM PT! LinkedIn: https://www.linkedin.com/events/makingllmsobservable-ai-opentel7357254715761446912/theater/ YouTube:

Hey folks! Almost all of us aren't big fans of YAML right? Here's a small trivia specifically designed to test your YAML knowledge. We are also giving very cool swags for top scorers! So try it out and flaunt your scores! 🤓 Hope you don't lose your sanity playing this! 😉 ohyaml.wtf

Hello <!here>, We're launching our ambassador program soon and want your thoughts on the name before we make it official. The program will recognize active community members who contribute through content creation, helping others, speaking at events, and spreading the word about SigNoz. They will receive exclusive swag boxes and other perks. Program name: SigNoz Nightwatchers Individual title: Nightwatcher So someone would add "Nightwatcher at SigNoz" to their LinkedIn profile, Twitter bio, etc. The "Nightwatchers" theme plays on the idea of always being vigilant about system observability - keeping watch over applications even in the "dark" hours when issues might strike. What do you think? • Does the name resonate with you? • Any immediate reactions or concerns? We've started the prep work, but need your feedback before we finalize everything 🙌