Search logs without losing the execution context.
Logs are where teams often see symptoms first. CloudGrid keeps OTLP log records searchable while preserving severity, service, attributes, trace id, and span id, so a useful line can lead directly to the trace that explains it.
Logs become decision evidence when they stay connected.
Enterprise operators need fast search, but they also need confidence that a log line belongs to the right customer, project, service, and request. CloudGrid keeps those relationships explicit instead of turning logs into a detached text archive.
From text symptom to operational follow-through.
The log page is intentionally not a separate world. It is a fast entry point into the same project record that also owns traces, metrics, dashboards, alerts, and evaluations.
Operators need fast search over message bodies, severity, services, attributes, and time ranges when a system starts behaving differently.
A useful log entry should not become a dead end. Trace and span ids make the next click a pivot into the actual request path.
When a log pattern matters repeatedly, it can become an alert condition, a dashboard widget, or part of an incident review.
What CloudGrid provides
- OTLP HTTP ingest on /v1/logs with JSON and protobuf encodings.
- Severity-aware table for INFO, WARN, ERROR, DEBUG, and related log severity values.
- Full-text search over message bodies plus structured filters on attributes.
- Trace-id and span-id columns that link a log row back to its trace.
- Project-scoped queries so users only see logs for projects they belong to.
Operating model
- Redaction policies are owned by your collector pipeline; CloudGrid stores what is sent.
- Log match and log count alerting live in the project alerting workspace.
- Metrics remain the time-series path for aggregation questions.
- Long-term archival belongs behind alternative storage adapters.
Where connected logs help.
Logs are often the first human-readable signal. CloudGrid keeps them useful after the first search.
Customer-facing incidents can move from a text symptom to the request that produced it.
Teams can keep structured filters and project scoping instead of searching a shared log bucket.
Redaction remains in the customer collector pipeline, while CloudGrid stores and queries what was intentionally sent.
