About Relay

Relay is a log forwarding service built for teams that need their log data in more than one place. Instead of locking your logs into a single vendor, Relay sits between your applications and your storage, routing every event wherever it needs to go.

Most teams start with one logging destination and end up with three or four. An Elasticsearch cluster for search, S3 for long-term archival, a SIEM for security, and maybe a webhook that fires on errors. Stitching those integrations together with custom code is fragile and time-consuming. Relay replaces that glue.

The service accepts logs over HTTP, syslog, or raw TCP, applies the filtering and transformation rules you define in YAML, and forwards the result to as many destinations as you configure. Delivery is at-least-once with automatic retries. Failed deliveries are buffered and replayed when the destination recovers.

Relay operates in 12 regions across North America, Europe, and Asia-Pacific. All data is encrypted in transit and at rest. Region-locked pipelines ensure your logs never leave the geography you specify.

How we think about reliability

A log forwarder is only useful if it is always running. Relay is deployed across multiple availability zones in every region. Each component — ingestion, processing, forwarding — scales independently and fails independently. If a single forwarding worker goes down, the remaining workers absorb its share within seconds.

We publish real-time system status and a full incident history on our status page. Every incident gets a public post-mortem within 48 hours.