OGM Inspector

2 open findingsin stage

Grouped under the component believed to be causing them.

Broken route — nothing is listening on port 32229

critical

Broken route — nothing is listening on port 32229

first seen 26h ago

nginx is configured to forward traffic to 127.0.0.1:32229, but no process on this host is bound to that port. Requests to stage.samskriti.ogmworld.com will return 502 Bad Gateway.

How to resolve

Start the service that is supposed to own port 32229, or remove the stale nginx route if the application has been retired. Check 'docker ps -a' and 'systemctl list-units --state=failed' on this host — a container or unit that exited is the usual cause.

ogm:stage:stage-kvm2:port:127.0.0.1-32229

Stage — KVM2 is cut off from every shared datastore

critical

Stage — KVM2 is cut off from every shared datastore

first seen 42m ago

chromadb, elasticsearch, mongodb, neo4j, nfs, postgresql, qdrant, redis are healthy on ogm-datacenter, but Stage — KVM2 cannot open a TCP connection to them (ports 6379, 6333, 2049, 9200, 8001, 5433, 27017, 7687). The services are up; the network path is not. Every expected datastore is blocked, which points at this host being absent from the datacenter firewall allow-list rather than at any individual service.

How to resolve

This is a network path problem, not a service problem. ogm-datacenter publishes these ports through Docker, which bypasses plain 'ufw allow' rules — the allow-list lives in the DOCKER-USER chain, maintained by scripts/setup-firewall.sh in the ogm-datacenter repository. Add this host's public IP to the ALLOWED_SOURCES array in that script and re-run it, then re-collect to confirm.

ogm:stage:stage-kvm2:host:stage-kvm2blocks 8 components