removed an integration from this channel: linen.dev

Hi! Could someone take a look at my PR?▎ • PR: github.com/openobserve/openobserve/pull/14825 • Issue: github.com/openobserve/openobserve/issues/14241 • RCA: github.com/openobserve/openobserve/issues/14241#… It fixes schema rows being saved with start_dt = 0, which renders a 3-segment key that db:schema:cache() asserts on — the node then panics on every restart and crash-loops. Adds a migration that repairs stores already holding such a row.

🚀 From days to seconds: how Decklar scaled observability on just 4 machines Shailesh Mangal, CTO at Decklar, explains how troubleshooting that used to take days (sometimes a week) now takes seconds with OpenObserve, without adding more hardware as they grow. ⚡ Quick highlights: • Scaling on the same 4-machine cluster • Alerts in place of manual querying • All monitoring data in one place, which means fewer incidents ⏱️ About 2 mins 👉 💬 What's the hardest part of scaling observability for your team? Share in the thread 🧵

Hi! Could someone take a look at my PR? PR: github.com/openobserve/openobserve/pull/14599 Issue: github.com/openobserve/openobserve/issues/14408 RCA: github.com/openobserve/openobserve/issues/14408#… It fixes OTLP/JSON ingestion dropping the entire batch when any record carries a fractional number ({"doubleValue": 1.5} → 400), and metrics with asDouble returning 200 but storing nothing.

Setting up CockroachDB monitoring with OpenObserve? Check out our latest community-contributed blog to get started. Read it here: openobserve.ai/blog/monitor-cockroachdb Shoutout to @anurag for creating this guide and adding 3 new Monitoring dashboards [CockroachDB, Nginx and Kafka] to the Dashboard Repository. Thanks for sharing with the community! Got something to share with the community? We'd love to see your dashboards, blogs, and OpenObserve projects. Keep the contributions coming!

<!here> Join our September 30 webinar featuring engineering experts from Apple, SLB, and Cars Commerce as they share production war stories and lessons you won’t find in a playbook. Learn what they wish they’d put in place earlier, what they’d avoid, and what they’ve had to rethink as complexity grew. Practical lessons you can apply now, whether you’re building the foundation or already operating at scale. 👉 Join live for Q&A, or register to get the recording if you can’t attend: openobserve.ai/webinars-videos/alert-fatigue-noise-slos-incidents-production

Quick thank you to everyone who showed up for the AI Observability launch on Product Hunt yesterday. We finished #2 product of the day 💪 If you haven't tried AI Observability yet, it's live now. Point your agent traces at OpenObserve and tell us what you think. And if you left a comment or an upvote, thank you. This community is the reason we get to build this.

<!here> help us reach the #1 spot on Product Hunt. Every interaction counts. Drop an upvote or feedback! producthunt.com/posts/ai-observability-by-openobserve

Hello everyone, upvoted!

<!here> We’re live on PH, and your upvotes, comments, questions, and conversations are what keep the momentum going. If you haven’t stopped by yet, head over, show some love, and help us get the launch in front of more people. Every interaction counts. producthunt.com/posts/ai-observability-by-openobserve

<!here> We’re currently sitting at #1 on Product Hunt and the momentum is real. Help us keep it going! If you’ve been meaning to check out AI Observability, now’s a great time drop your questions, share your thoughts, or just come say hi on the launch page. Every interaction helps us stay visible and keep the momentum going. 👉 Come hang out with us on Product Hunt: producthunt.com/posts/ai-observability-by-openobserve

🚨AI Observability is live on Product Hunt today! We’ve been working on bringing LLM + agent observability together with logs, metrics, and traces, with visibility into costs, tokens, latency, and evaluations. Would love for the OpenObserve community to check it out, share your feedback, and drop any questions directly on the launch page: producthunt.com/posts/ai-observability-by-openobserve

Hi team, we have identified two issues: 1. Querying and downloading large volumes of data When running queries for data from one month prior, where the Delta Scan Size is around 469.63 GB, there is an issue when navigating beyond the first 50 results. When I try to load the next 50 records, the entire interface freezes and eventually crashes, which means I’m unable to retrieve the remaining data points for the one-month period. A similar issue occurs when trying to download a larger number of records. For example, when attempting to download 1,000 records, the download appears to time out and fails completely. It is understandable if the query or download takes some time due to the volume of data, but it should ideally continue processing rather than freezing or timing out halfway through. 2. Unnecessary data being downloaded when querying historical logs When searching for logs from a specific historical date, for example 7 August, it appears that the system is still downloading data starting from the current/latest time rather than only retrieving the relevant data for the requested date. This results in a significant amount of unnecessary data being downloaded and increases the time required to complete the query. Considering how the data is organised in S3, it seems possible to identify and retrieve only the relevant data partitions/files for the requested date range instead of downloading data from the current time backwards. Could you please look into both of these issues?

Hi team, I want to work on this issue github.com/openobserve/openobserve/issues/6582 and added my approach few days back. Just posting here for visibility since it's an old issue Thanks

Hi team, I opened a PR to fix alert and anomaly scheduler drift. The fix keeps interval-based schedules anchored to their previous scheduled time and skips missed intervals after delays. I reproduced the issue locally and added focused tests for normal and delayed scheduling. PR: github.com/openobserve/openobserve/pull/14084

Your Kubernetes bill grows. Your observability bill grows faster. 😵‍💫 That’s exactly what happened at Uno.ai. As their AI infrastructure grew across EKS, self-managed Kubernetes, and private cloud, observability started getting way too expensive. Their solution? Move to OpenObserve and stop letting observability costs grow unchecked. Shashank Tiwari, CEO & Co-founder of Uno.ai, talks about the move, the migration, and how they’re keeping full visibility without the “wait… why is observability this expensive?” moment. 🎥 Watch the story: What’s been your biggest observability headache lately ? cost, complexity, or actually finding the signal? 👀

Hi team , I opened PR #14031 to prevent a node from being published

Online

before its gRPC listener is ready. I confirmed the startup ordering could signal initialization before Tonic bound the gRPC socket, leaving a brief window for distributed queries to select an unreachable node. The fix binds the listener first, then reports readiness, and adds coverage for successful readiness and bind failures. CI is running.

I wonder if Vortex is the default. From the blog post it looks like the only way is to set an env variable which for some reason isn't documented in the official docs.

Hi team, I reproduced issue #13926 and opened PR #13937. The problem was that mixed-case metric names were being changed to lowercase while the WAL query moved between services. The fix keeps the original name unchanged during that step. I also added regression tests. Requesting for a review on PR.

Hi team, I opened PR #13924 to fix #13908. The fix keeps inferred services with the same name separate by preserving their type and system identity across the Service Catalog, side panel queries, and trace navigation. I verified it against a running OpenObserve instance using real OTLP data. Two

sso

services now appear as separate database and external rows, and repeated ingestion updates only the correct row. The production build, lint, formatting, type checks, and affected trace tests passed. The full test suite had one unrelated timeout, which passed when rerun independently. Feedback on the PR is welcome.

Hi Team; Do you have any hardening/security guide for the agent when deployed on VMs?

Does O2 expose Baggage at all in traces? Would that even be a good idea?

Hello folks , I had setup OpenObserve , but all the issues seems to be pre-assigned on github . Can someone point at some good issues to work on ?

Real time alerts are not working after version update to v0.92.0. Similar issue on v0.80.3

Self-hosted versions

How is the cost of the Business Edition determined?

👋 Hi everyone!

Hi i am new to openobserve just wandering to create an alert destination as Outlook email, suggestions will be appreciated

<!here> What do you guys think about adding this feature?

At 3:40 am central time (USA) in the dashboard section of the configured alert there was a spike in memory usage for the pod vault-2 and the alert should have triggered around that time on slack channel. but the alert got triggered at 5:03 am . So there is a delay in delivery of slack channel alert and Alert according to dashboard by 1 hr 20 min . Also have attached the json config for the alert. Would like to achieve the slack channel alert close to the Alert according to dashboard: 1st screenshot: Slack channel alert : 2nd screenshot is : Alert according to dashboard: