loki-logger

loki-logger

Builds a JSON access-log entry for each request/response and ships accumulated batches to a Grafana Loki push endpoint. Each batch is grouped into a single Loki stream carrying the configured labels, with one [nanosecond_timestamp, json_line] value per entry, and POSTed to <endpoint>/loki/api/v1/push. Place this node in the response pipeline, after the upstream node.

Configuration

Key	Type	Default	Description
endpoint_addrs	array	— (required)	Loki base addresses (e.g. http://loki:3100); one is chosen per flush. A single-string endpoint is also accepted.
endpoint_uri	string	/loki/api/v1/push	Push path appended to each base address.
tenant_id	string	fake	Sent as the X-Scope-OrgID header.
log_labels	object	{job: featherbit}	Loki stream labels. labels is accepted as an alias.
headers	object	{}	Extra request headers.
ssl_verify	bool	false	Verify TLS certificates.
timeout	integer (ms)	3000	Whole-call deadline per flush.
log_format	object	—	Custom flat entry of name -> "$var template".
include_req_body	bool	false	Add the request body to the default entry.
include_resp_body	bool	false	Add the response body to the default entry.
batch_max_size	integer	1000	Flush when the buffer reaches this many entries.
inactive_timeout	integer (s)	5	Flush after this idle period.
buffer_duration	integer (s)	60	Flush when the oldest buffered entry is this old.
max_retry_count	integer	0	Retries after a failed flush before dropping the batch.
retry_delay	integer (s)	1	Delay between retries.
max_pending_entries	integer	10000	Queue capacity; entries are dropped with a warning when full.

- id: access-log
  type: loki-logger
  config:
    endpoint_addrs: [http://loki:3100]
    tenant_id: my-org
    log_labels:
      job: featherbit
      env: prod
    batch_max_size: 1000

Behavior

The node is a pure passthrough: it never modifies the context and never fails, so only its success port is ever taken. push is fire-and-forget and never blocks the request path — when the queue is full, entries are dropped with a tracing::warn!. Delivery, batching, timing, and retries all run on a background task. Note: label values are static (no per-request $var resolution), so a batch forms a single stream.