We run HyperDX self-hosted on EC2/ECS with the OTel Collector — no Kubernetes anywhere.
The Infrastructure tab never appears in our log side panel. Looking at infraCorrelations.ts, the tab is gated on k8s.pod.uid / k8s.node.name: when neither resource attribute is present the tab isn't added at all, so there's nothing to see rather than an empty panel. Setting a "Correlated Metric Source" on the log source doesn't change that — it only picks which metric source would be queried.
Our logs do carry host.name, and we can ship host metrics over OTLP, so the data side is there — it's just not something the current correlation list looks for.
Are there plans to support host-level correlation (e.g. host.name with system.* metrics), or is Kubernetes the intended scope for this feature?
Asking before we invest in workarounds. If it's something you'd take, we're happy to help — including testing a branch against a real non-k8s deployment.
Environment: HyperDX 2.33.0 self-hosted, external ClickHouse, EC2/ECS, no Kubernetes.
We run HyperDX self-hosted on EC2/ECS with the OTel Collector — no Kubernetes anywhere.
The Infrastructure tab never appears in our log side panel. Looking at
infraCorrelations.ts, the tab is gated onk8s.pod.uid/k8s.node.name: when neither resource attribute is present the tab isn't added at all, so there's nothing to see rather than an empty panel. Setting a "Correlated Metric Source" on the log source doesn't change that — it only picks which metric source would be queried.Our logs do carry
host.name, and we can ship host metrics over OTLP, so the data side is there — it's just not something the current correlation list looks for.Are there plans to support host-level correlation (e.g.
host.namewithsystem.*metrics), or is Kubernetes the intended scope for this feature?Asking before we invest in workarounds. If it's something you'd take, we're happy to help — including testing a branch against a real non-k8s deployment.
Environment: HyperDX 2.33.0 self-hosted, external ClickHouse, EC2/ECS, no Kubernetes.