This scenario adds support for metrics use cases to the LogStream demo. We open a metrics input on dogstatsd and graphite protocols. We add pipelines for processing that data. We add pipelines for processing weblog data, aggregating it into metric events and then outputting to Influxdb. This scenario also adds InfluxDB as a datastore and Grafana as a visualization engine.
Description | URL | Username | Password |
---|---|---|---|
Cribl | http://localhost:9000 | admin | cribldemo |
Grafana | http://localhost:8200 | admin | cribldemo |
Description | URL | Username | Password |
---|---|---|---|
MTR Pipeline | http://localhost:9000/pipelines/mtr | admin | cribldemo |
Logs To Metrics Pipeline | http://localhost:9000/pipelines/logs_to_metrics | admin | cribldemo |
In Grafana, we ship a dashboard which shows the output of MTR which is kind of a combination of ping and traceroute. We have mtr
output JSON, and use nc
to send the data to a Cribl TCPJSON input. The MTR Pipeline takes the JSON from mtr
and explodes it into an event for every hop along the network path. It then resolves the host IP from DNS and publishes each of the hops as a metric to InfluxDB.
This shows a practical use case for taking data output by standard utilities and munging it easily into a format expected by another downstream datastore, in this case InfluxDB.