To send your Prometheus application metrics to a Logz.io Infrastructure Monitoring account, use remote write to connect to Logz.io as the endpoint. Your data is formatted as JSON documents by the Logz.io listener.
Multiple server environments: If you have multiple Prometheus server instances, you’ll have to add Logz.io as an endpoint for each instance.
Configuring Remote Write to Logz.io
Get your Logz.io Infrastructure Monitoring account information
Within Logz.io, look up the Listener host for your region (URL) and the Logz.io Metrics Account token.
You’ll find the correct Region and Listener URL for your region in the Regions and Listener Hosts table.
Look up your Prometheus Metrics account token in the Settings > Manage tokens > Data shipping tokens > Metrics tab of your Operations workspace. For more information, see Metrics shipping token.
Add a remote_write url
Configure your Prometheus yaml file or use a Helm chart:
To configure a Prometheus yaml file
Add the following parameters to your Prometheus yaml file:
|external_labels||Parameters to tag the metrics from this specific Prometheus server.|
|p8s_logzio_name||Use the value of the parameter
|remote_write||The remote write section configuration sets Logz.io as the endpoint for your Prometheus metrics data. Place this section at the same indentation level as the
|url||The Logz.io Listener URL for for your region, configured to use port 8052 for http traffic, or port 8053 for https traffic. For more details, see the Prometheus configuration file remote write reference.||Required|
|bearer_token||The Logz.io Prometheus Metrics account token.||Required|
global: external_labels: p8s_logzio_name: <labelvalue> remote_write: - url: https://<<LISTENER-HOST>>:8053 bearer_token: <<PROMETHEUS-METRICS-SHIPPING-TOKEN>> remote_timeout: 30s queue_config: batch_send_deadline: 5s #default = 5s max_shards: 10 #default = 1000 min_shards: 1 max_samples_per_send: 500 #default = 100 capacity: 10000 #default = 500
To configure a Helm chart
For kube-prometheus-stack Helm chart users:
Edit your chart
values.yaml file in the following sections:
- remote write:
remoteWrite: - url: https://<<LISTENER-HOST>>:8053 # The Logz.io Listener URL for your region, configured to use port **8052** for http traffic, or port **8053** for https traffic. bearerToken: <<PROMETHEUS-METRICS-SHIPPING-TOKEN>> # The Logz.io Prometheus metrics account token remoteTimeout: 30s queueConfig: batchSendDeadline: 5s #default = 5s maxShards: 10 #default = 1000 maxSamplesPerSend: 500 #default = 100 capacity: 10000 #default = 500
externalLabels: p8s_logzio_name: <labelvalue>
Verify the remote_write configuration
Run a query: If you are scraping Prometheus metrics, you can check that the remote_write configuration is working properly by doing one of the following and verifying that the result is greater than zero (n > 0) for the url:
Run a query on your local Prometheus interface for the metric
Check for the metric in the
/metricsendpoint on your Prometheus server.
Check via Metrics Explore: To verify that metrics are arriving to Logz,io:
Open Metrics Explore via the compass icon in the left menu bar.
Examine the Metrics dropdown list below the Explore heading in the upper left of the pane.
An empty list or the text no metrics indicates that the remote write configuration is not working properly.
Open Metrics Explore
Once you’ve verified that your data is available in Logz.io, explore your Prometheus metrics.
After your metrics are flowing, import your existing Prometheus and Grafana dashboards to Logz.io Infrastructure Monitoring as JSON files.
Reduce tagging: By default, all the metrics from your Prometheus server(s) are sent to Logz.io. To drop or send specific metrics, add Prometheus labeling before enabling the remote write, or as part of the remote write configuration. Learn more about Prometheus relabeling tricks here.
Metrics metadata dashboards: If you have both Prometheus & Grafana, you can activate a dashboard as part of the remote write configuration that will show you the queue size and how many metrics you’re sending. If your queue size increases, it might be necessary to open an additional channel. (currently in development)
Tune the remote write process: Learn more about Prometheus remote write tuning here.