Odigos

Outages start in the gaps others miss. See the hardest parts of your system and capture more context when it matters, so engineers and AI resolve issues faster.

[Odigos @odigosio]

We're a few weeks out from CNCF Observability Summit. If your day-one plan ends at "find somewhere to grab a drink after," we've got you covered! 🍻

Spots are limited.
RSVP: https://t.co/azNrqVyELp

#CloudNative #eBPF

[Odigos @odigosio]

Your Python service calls an LLM. Your trace shows one generic POST to OpenAI at 3.2s and nothing else.

No model. No tokens. No reason why.

We bundled 7 OTel GenAI instrumentations into the Odigos Python distro. The missing attributes show up automatically.

See them all here: https://t.co/q6sBrFhOR9

#OpenTelemetry #Observability

[Odigos @odigosio]

Heading to @CloudNativeFdn Observability Summit next month? Decompress with us after day one.

Happy hour on us, short walk from the convention center.

RSVP: https://t.co/i7QbvZ0auY

#Observability

[Odigos @odigosio]

Every observability setup has its dark matter: the services you can't see because they're too hard to instrument.

At CNCF Observability Summit (May 21–22, Minneapolis), Odigos CTO @edeNFed is taking the keynote stage to show how eBPF-based instrumentation makes that mass visible. Full observability across every service in the cluster, OpenTelemetry output to your existing backend, no code changes, <1% overhead.

But that's not all! We're a Diamond Sponsor at table T3. Come find us and catch Eden's keynote Thursday at 9:15 AM CDT.
https://t.co/5tZUK52WiP
👇
BTW 👉 20% discount code for attendees: OBS26SPODIG20. Register as an "Attendee" and apply the code on the Payment page.

#CNCF #Observability

[Odigos @odigosio]

When a Kafka consumer silently fails, how long does it take your team to notice?

Pods are healthy. Consumer lag looks normal. Nothing alerts.

Will Searle ran three of these scenarios and wrote up what it takes to catch them: https://t.co/r8YNlMACiZ

#Kafka #eBPF

[Odigos @odigosio]

From our recent livestream, where we talked about tracing Kubernetes control planes. Mike Dame talks about a scenario where a network policy blocked cross-namespace traffic and how Odigos picked up on it.

Watch 👉 https://t.co/9S5u1cYF0f

[Odigos @odigosio]

How much of your incident actually fits in a 1M-token context window? @edeNFed ran the numbers for a real-world scenario.

The answer is not great: https://t.co/ktWWnOU21C

#OpenTelemetry #Observability

[Odigos @odigosio]

Most observability setups have the same blind spot: compiled languages, third-party apps, legacy services, and latency-sensitive workloads.

These are exactly the places where the worst outages start.

💎 We're a Diamond Sponsor of @CloudNativeFdn CNCF Observability Summit (May 21–22, Minneapolis), and our CTO @edeNFed will break down how eBPF-based instrumentation with OpenTelemetry output closes these gaps, across every service in your cluster, with no code changes and <1% overhead.

Come see us if you will be there and catch Eden's session. If your postmortems keep ending with "we didn't have coverage there," this one is for you.
https://t.co/EDopbvTwhL

#CNCF #Observability #OpenTelemetry #eBPF #CloudNative #Kubernetes #O11ySummit

[Odigos @odigosio]

🧵 Your Kafka producer returns 200. The frontend shows a green checkmark.But one of your five consumers just silently failed. Pods are healthy. Consumer lag looks normal. Nothing alerts.

We walked through three scenarios where this happens. ↓

1/ Ghost SKU: A discontinued product is still in the catalog dropdown. Every order with that SKU passes the producer, passes four consumers, and quietly fails inventory validation with a 404.Without payload visibility in the trace, you'd never connect "inventory-checker failures" to a specific product ID.

2/ Phantom Timeout: Critical-priority orders aren't getting confirmation emails. The notification-sender pod looks fine. No restarts, normal memory/CPU.The trace shows an SMS gateway timeout (2000ms+, HTTP 503) that only triggers on critical-priority messages. You know which team to page and what to fix.

3/ Intermittent Infra: The audit-logger fails ~15% of writes. No pattern in the data. Same order succeeds once, fails the next.Comparing payloads of failing vs. succeeding traces: identical. That rules out every data issue and points straight to connection instability.

All three caught with eBPF-based auto-instrumentation. No SDK changes. No code modifications. Trace context propagated through Kafka headers automatically.

Full writeup https://t.co/FHRaA93Luq

#Kafka #OpenTelemetry

[Odigos @odigosio]

New Odigos UI Beta is here!

Sleek design and powerful observability tools—try it now: https://t.co/2Uv40nuBEe

[Odigos @odigosio]

We’re gearing up for a major update to the Odigos GUI! 📷 Expect improved navigation, a refreshed look, and features that make managing observability smoother than ever. We can’t wait to share this with you – stay tuned for the reveal!

[Odigos @odigosio]

💻 Discover the power of #eBPF! Dive into our latest article on how eBPF transforms #Observability with real-time, in-kernel insights—no code changes, minimal overhead. Perfect for cloud-native, high-performance systems!

Read more 👉 https://t.co/bCvnnfLBjE

[Odigos @odigosio]

🚀 Thinking about migrating to OpenTelemetry but not sure where to start? We've got you covered with our step-by-step migration guide!

https://t.co/xU4NQ9WLSo

[Odigos @odigosio]

Join @Odigosio at #KubeCon + #CloudNativeCon!
Discover zero-code instrumentation for faster setup, &lt;1% performance overhead, and support for compiled languages!

Nov 12-15, 2024 | Salt Palace Convention Center
🔗 RSVP today! https://t.co/R5cvxWFEcX

[Odigos @odigosio]

With @Odigosio, you get deep insights without code changes, low overhead, and real-time security monitoring.

Learn how Odigos leverages eBPF to optimize performance and visibility. 🔍
https://t.co/SAKQ9FAzfN

#Observability #CloudNative

[Odigos @odigosio]

Watch the live session recording: "The Future of eBPF" with @edeNFed!

Learn how eBPF is transforming #Kubernetes observability &amp; security with real-time monitoring and minimal overhead: https://t.co/vXEZeuO3QO

[Odigos @odigosio]

Struggling with #Observability in complex systems?

Check out our latest article on key metrics &amp; tools every #DevOps &amp; #SRE team should use for distributed tracing!

Learn how @odigosio makes it seamless—no code changes needed! 👇

https://t.co/icOqJ2uJqv

[Odigos @odigosio]

Learn about essential observability like latency &amp; error rates and the top tools—Prometheus, Grafana, Jaeger, OpenTelemetry, and Odigos— for a robust observability strategy.

Read more: https://t.co/syR38JB7d4 🔗

[Odigos @odigosio]

Curious how Odigos started and evolved?
Check out this TechCrunch article that shares our journey, from simplifying distributed tracing to helping enterprises eliminate system errors and latency.

[TechCrunch @TechCrunch]

YC alum Odigos aims to help enterprises find errors and stamp out latency in their systems https://t.co/nU5RcVuwrD

[Odigos @odigosio]

Thrilled to announce we've raised $13M led by Venture Guides with support from Salesforce Ventures, Mango Capital, Firestreak Ventures, and industry leaders like Y Combinator, Martin Mao, Christine Yen, Ben Sigelman, and Yuri Shkuro! 🚀

Stay tuned!

https://t.co/1FTGBou9UT