# `Slap.Streams.Telemetry`
[🔗](https://github.com/mjrusso/slap/blob/slap_streams-v0.3.0/slap_streams/lib/slap/streams/telemetry.ex#L1)

Telemetry events, all under `[:slap, :streams, ...]`. `Slap.Streams.Metrics` turns
them, and `slap_cluster`'s, into Prometheus metrics.

  * `[:append, :acknowledged]` - an append was acknowledged (durable).
    Measurements: `:duration` (native time units, from the stream server
    accepting it to the reply), `:bytes`. Metadata: `:shard`, `:result`
    (`:appended` or `:closed`).
  * `[:append, :rejected]` - an append was refused for backpressure.
    Measurements: `:bytes`. Metadata: `:shard`, `:reason` (`:overloaded`).
  * `[:stream_server, :write_failed]` - a write failed; the server stopped
    and failed its in-flight requests. Metadata: `:shard`.
  * `[:wait, :registered]` - a read with `:wait` (long poll or SSE) found
    no data and is waiting. Any append after this event wakes it.
    Metadata: `:shard`, `:path`, `:offset` (the offset it waits for).
  * `[:shard, :load]` - every `:load_interval`, per shard: the
    measurements of the shard's load. Metadata: `:shard`.
  * `[:http, :request]` - an HTTP response was sent. Measurements:
    `:duration`. Metadata: `:method`, `:status`.

The acknowledgement and wait events have no matching `:start` or `:stop`
events.

---

*Consult [api-reference.md](api-reference.md) for complete listing*