🔗 Repositorio oficial y código:
OneUptime/oneuptime
📌 Guarda este post para tu próxima auditoría de costos SRE.
1 view
I built an open-source alternative to Datadog, PagerDuty, and Statuspage
I’m building OneUptime (oneuptime/oneuptime) an open-source observability platform that brings monitoring, alerting, incident management, status pages, logs, metrics, and traces into one place.
Self-host it on your own infrastructure or use our cloud. Free and open source under Apache 2.0.
Our goal is simple: help teams build more reliable software and avoid paying the observability tax to Datadog / NewRelic / etc.
Let me know what you think!
9 comments
Kubernetes telemetry that stays on your own infrastructure
A recurring theme in k8s monitoring discussions: you want rich cluster telemetry, but you do not want to ship all of it to a SaaS that bills per host. You want the data in a place you control.
I work on OneUptime, an open source observability platform (Apache 2.0). There is a small Kubernetes agent chart that collects cluster telemetry and sends it to the main app over OpenTelemetry. The main app is one Helm chart. So the whole setup is two charts on your infrastructure, and the data stays with you.
It covers uptime checks, status pages, on-call paging, incident management with postmortems, plus log, metrics, and APM views. It is younger than the big commercial observability stacks, and some features are in a separate enterprise edition. Read the code before you trust it.
GitHub: OneUptime/oneuptime
Site: oneuptime.com
What is the one k8s signal you consider non-negotiable in a monitoring setup?
6 comments
On-call and incident management without a per-host observability bill
There is a quiet cost to commercial observability: the bill scales with your infrastructure. More hosts, more metrics, more money, until finance starts asking questions. And on-call and incident tooling is often a separate line item on top.
I work on OneUptime, an open source observability platform (Apache 2.0). The part I think DevOps teams underestimate is that on-call scheduling, paging, incident management, and postmortems live in the same app as uptime and status pages. One self-hosted system: a Helm chart for the app, a small chart for the Kubernetes agent, data in over OpenTelemetry, telemetry on your infrastructure.
It is not a drop-in for a five-year Datadog or Grafana estate, and it is younger than the commercial platforms. A few features live in a separate enterprise edition. I would judge it by the repo, not a sales call.
GitHub: OneUptime/oneuptime
Site: oneuptime.com
For people who have run on-call: what is the single most annoying part of your current incident workflow?
2 comments
Mentions
🔗 Repositorio oficial y código: OneUptime/oneuptime 📌 Guarda este post para tu próxima auditoría de costos SRE.
I built an open-source alternative to Datadog, PagerDuty, and Statuspage I’m building OneUptime (oneuptime/oneuptime) an open-source observability platform that brings monitoring, alerting, incident management, status pages, logs, metrics, and traces into one place. Self-host it on your own infrastructure or use our cloud. Free and open source under Apache 2.0. Our goal is simple: help teams build more reliable software and avoid paying the observability tax to Datadog / NewRelic / etc. Let me know what you think!
Kubernetes telemetry that stays on your own infrastructure A recurring theme in k8s monitoring discussions: you want rich cluster telemetry, but you do not want to ship all of it to a SaaS that bills per host. You want the data in a place you control. I work on OneUptime, an open source observability platform (Apache 2.0). There is a small Kubernetes agent chart that collects cluster telemetry and sends it to the main app over OpenTelemetry. The main app is one Helm chart. So the whole setup is two charts on your infrastructure, and the data stays with you. It covers uptime checks, status pages, on-call paging, incident management with postmortems, plus log, metrics, and APM views. It is younger than the big commercial observability stacks, and some features are in a separate enterprise edition. Read the code before you trust it. GitHub: OneUptime/oneuptime Site: oneuptime.com What is the one k8s signal you consider non-negotiable in a monitoring setup?
On-call and incident management without a per-host observability bill There is a quiet cost to commercial observability: the bill scales with your infrastructure. More hosts, more metrics, more money, until finance starts asking questions. And on-call and incident tooling is often a separate line item on top. I work on OneUptime, an open source observability platform (Apache 2.0). The part I think DevOps teams underestimate is that on-call scheduling, paging, incident management, and postmortems live in the same app as uptime and status pages. One self-hosted system: a Helm chart for the app, a small chart for the Kubernetes agent, data in over OpenTelemetry, telemetry on your infrastructure. It is not a drop-in for a five-year Datadog or Grafana estate, and it is younger than the commercial platforms. A few features live in a separate enterprise edition. I would judge it by the repo, not a sales call. GitHub: OneUptime/oneuptime Site: oneuptime.com For people who have run on-call: what is the single most annoying part of your current incident workflow?