Fluentd setup

VictoriaLogs supports given below Fluentd outputs:

Loki
HTTP JSON

Loki #

Specify loki output section in the fluentd.conf for sending the collected logs to VictoriaLogs:

<match **>
  @type loki
  url "http://localhost:9428/insert"
  <buffer>
    flush_interval 10s
    flush_at_shutdown true
  </buffer>
  custom_headers {"VL-Msg-Field": "log", "VL-Time-Field": "time", "VL-Stream-Fields": "path"}
  buffer_chunk_limit 1m
</match>

HTTP #

Specify http output section in the fluentd.conf for sending the collected logs to VictoriaLogs:

<match **>
  @type http
  endpoint "http://localhost:9428/insert/jsonline"
  headers {"VL-Msg-Field": "log", "VL-Time-Field": "time", "VL-Stream-Fields": "path"}
</match>

Substitute the host (localhost) and port (9428) with the real TCP address of VictoriaLogs.

See these docs for details on the query args specified in the endpoint.

It is recommended verifying whether the initial setup generates the needed log fields and uses the correct stream fields. This can be done by specifying debug parameter in the endpoint and inspecting VictoriaLogs logs then:

<match **>
  @type http
  endpoint "http://localhost:9428/insert/jsonline&debug=1"
  headers {"VL-Msg-Field": "log", "VL-Time-Field": "time", "VL-Stream-Fields": "path"}
</match>

If some log fields must be skipped during data ingestion, then they can be put into ignore_fields parameter. For example, the following config instructs VictoriaLogs to ignore log.offset and event.original fields in the ingested logs:

<match **>
  @type http
  endpoint "http://localhost:9428/insert/jsonline&ignore_fields=log.offset,event.original"
  headers {"VL-Msg-Field": "log", "VL-Time-Field": "time", "VL-Stream-Fields": "path"}
</match>

If the Fluentd sends logs to VictoriaLogs in another datacenter, then it may be useful enabling data compression via compress gzip option. This usually allows saving network bandwidth and costs by up to 5 times:

<match **>
  @type http
  endpoint "http://localhost:9428/insert/jsonline&ignore_fields=log.offset,event.original"
  headers {"VL-Msg-Field": "log", "VL-Time-Field": "time", "VL-Stream-Fields": "path"}
  compress gzip
</match>

By default, the ingested logs are stored in the (AccountID=0, ProjectID=0) tenant. If you need storing logs in other tenant, then specify the needed tenant via header options. For example, the following fluentd.conf config instructs Fluentd to store the data to (AccountID=12, ProjectID=34) tenant:

<match **>
  @type http
  endpoint "http://localhost:9428/insert/jsonline"
  headers {"VL-Msg-Field": "log", "VL-Time-Field": "time", "VL-Stream-Fields": "path"}
  header AccountID 12
  header ProjectID 23
</match>

Fluentd setup #

Loki #

HTTP #