Exporting engine reference

Welcome to the exporting engine reference guide. This guide contains comprehensive information about enabling, configuring, and monitoring Netdata's exporting engine, which allows you to send metrics to external time-series databases.

For a quick introduction to the exporting engine's features, read our doc on exporting metrics to time-series databases, or jump in to enabling a connector.

The exporting engine has a modular structure and supports metric exporting via multiple exporting connector instances at the same time. You can have different update intervals and filters configured for every exporting connector instance.

The exporting engine has its own configuration file exporting.conf. The configuration is almost similar to the deprecated backends system. The most important difference is that type of a connector should be specified in a section name before a colon and an instance name after the colon. Also, you can't use host tags anymore. Set your labels using the [host labels] section in netdata.conf.

Since Netdata collects thousands of metrics per server per second, which would easily congest any database server when several Netdata servers are sending data to it, Netdata allows sending metrics at a lower frequency, by resampling them.

So, although Netdata collects metrics every second, it can send to the external database servers averages or sums every X seconds (though, it can send them per second if you need it to).

Features

  1. The exporting engine uses a number of connectors to send Netdata metrics to external time-series databases. See our list of supported databases for information on which connector to enable and configure for your database of choice.

  2. Netdata can filter metrics (at the chart level), to send only a subset of the collected metrics.

  3. Netdata supports three modes of operation for all exporting connectors:

    • as-collected sends to external databases the metrics as they are collected, in the units they are collected. So, counters are sent as counters and gauges are sent as gauges, much like all data collectors do. For example, to calculate CPU utilization in this format, you need to know how to convert kernel ticks to percentage.

    • average sends to external databases normalized metrics from the Netdata database. In this mode, all metrics are sent as gauges, in the units Netdata uses. This abstracts data collection and simplifies visualization, but you will not be able to copy and paste queries from other sources to convert units. For example, CPU utilization percentage is calculated by Netdata, so Netdata will convert ticks to percentage and send the average percentage to the external database.

    • sum or volume: the sum of the interpolated values shown on the Netdata graphs is sent to the external database. So, if Netdata is configured to send data to the database every 10 seconds, the sum of the 10 values shown on the Netdata charts will be used.

    Time-series databases suggest to collect the raw values (as-collected). If you plan to invest on building your monitoring around a time-series database and you already know (or you will invest in learning) how to convert units and normalize the metrics in Grafana or other visualization tools, we suggest to use as-collected.

    If, on the other hand, you just need long term archiving of Netdata metrics and you plan to mainly work with Netdata, we suggest to use average. It decouples visualization from data collection, so it will generally be a lot simpler. Furthermore, if you use average, the charts shown in the external service will match exactly what you see in Netdata, which is not necessarily true for the other modes of operation.

  4. This code is smart enough, not to slow down Netdata, independently of the speed of the external database server. You should keep in mind though that many exporting connector instances can consume a lot of CPU resources if they run their batches at the same time. You can set different update intervals for every exporting connector instance, but even in that case they can occasionally synchronize their batches for a moment.

Configuration

Here are the configruation blocks for every supported connector. Your current exporting.conf file may look a little different.

You can configure each connector individually using the available options. The [graphite:my_graphite_instance] block contains examples of some of these additional options in action.

[exporting:global]
enabled = yes
send configured labels = no
send automatic labels = no
update every = 10
[prometheus:exporter]
send names instead of ids = yes
send configured labels = yes
end automatic labels = no
send charts matching = *
send hosts matching = localhost *
prefix = netdata
[graphite:my_graphite_instance]
enabled = yes
destination = localhost:2003
data source = average
prefix = Netdata
hostname = my-name
update every = 10
buffer on failures = 10
timeout ms = 20000
send charts matching = *
send hosts matching = localhost *
send names instead of ids = yes
send configured labels = yes
send automatic labels = yes
[prometheus_remote_write:my_prometheus_remote_write_instance]
enabled = yes
destination = localhost
remote write URL path = /receive
[kinesis:my_kinesis_instance]
enabled = yes
destination = us-east-1
stream name = netdata
aws_access_key_id = my_access_key_id
aws_secret_access_key = my_aws_secret_access_key
[pubsub:my_pubsub_instance]
enabled = yes
destination = pubsub.googleapis.com
credentials file = /etc/netdata/pubsub_credentials.json
project id = my_project
topic id = my_topic
[mongodb:my_mongodb_instance]
enabled = yes
destination = localhost
database = my_database
collection = my_collection
[json:my_json_instance]
enabled = yes
destination = localhost:5448
[opentsdb:my_opentsdb_plaintext_instance]
enabled = yes
destination = localhost:4242
[opentsdb:http:my_opentsdb_http_instance]
enabled = yes
destination = localhost:4242
[opentsdb:https:my_opentsdb_https_instance]
enabled = yes
destination = localhost:8082

Sections

  • [exporting:global] is a section where you can set your defaults for all exporting connectors
  • [prometheus:exporter] defines settings for Prometheus exporter API queries (e.g.: http://NODE:19999/api/v1/allmetrics?format=prometheus&help=yes&source=as-collected).
  • [<type>:<name>] keeps settings for a particular exporting connector instance, where:
  • type selects the exporting connector type: graphite | opentsdb:telnet | opentsdb:http | opentsdb:https | prometheus_remote_write | json | kinesis | pubsub | mongodb
  • name can be arbitrary instance name you chose.

Options

Configure individual connectors and override any global settings with the following options.

  • enabled = yes | no, enables or disables an exporting connector instance

  • destination = host1 host2 host3 ..., accepts a space separated list of hostnames, IPs (IPv4 and IPv6) and ports to connect to. Netdata will use the first available to send the metrics.

    The format of each item in this list, is: [PROTOCOL:]IP[:PORT].

    PROTOCOL can be udp or tcp. tcp is the default and only supported by the current exporting engine.

    IP can be XX.XX.XX.XX (IPv4), or [XX:XX...XX:XX] (IPv6). For IPv6 you can to enclose the IP in [] to separate it from the port.

    PORT can be a number of a service name. If omitted, the default port for the exporting connector will be used (graphite = 2003, opentsdb = 4242).

    Example IPv4:

destination = 10.11.14.2:4242 10.11.14.3:4242 10.11.14.4:4242

Example IPv6 and IPv4 together:

destination = [ffff:...:0001]:2003 10.11.12.1:2003

When multiple servers are defined, Netdata will try the next one when the previous one fails.

Netdata also ships nc-exporting.sh, a script that can be used as a fallback exporting connector to save the metrics to disk and push them to the time-series database when it becomes available again. It can also be used to monitor / trace / debug the metrics Netdata generates.

For the Kinesis exporting connector destination should be set to an AWS region (for example, us-east-1).

For the MongoDB exporting connector destination should be set to a MongoDB URI.

For the Pub/Sub exporting connector destination can be set to a specific service endpoint.

  • data source = as collected, or data source = average, or data source = sum, selects the kind of data that will be sent to the external database.

  • hostname = my-name, is the hostname to be used for sending data to the external database server. By default this is [global].hostname.

  • prefix = Netdata, is the prefix to add to all metrics.

  • update every = 10, is the number of seconds between sending data to the external datanase. Netdata will add some randomness to this number, to prevent stressing the external server when many Netdata servers send data to the same database. This randomness does not affect the quality of the data, only the time they are sent.

  • buffer on failures = 10, is the number of iterations (each iteration is update every seconds) to buffer data, when the external database server is not available. If the server fails to receive the data after that many failures, data loss on the connector instance is expected (Netdata will also log it).

  • timeout ms = 20000, is the timeout in milliseconds to wait for the external database server to process the data. By default this is 2 * update_every * 1000.

  • send hosts matching = localhost * includes one or more space separated patterns, using * as wildcard (any number of times within each pattern). The patterns are checked against the hostname (the localhost is always checked as localhost), allowing us to filter which hosts will be sent to the external database when this Netdata is a central Netdata aggregating multiple hosts. A pattern starting with ! gives a negative match. So to match all hosts named *db* except hosts containing *child*, use !*child* *db* (so, the order is important: the first pattern matching the hostname will be used - positive or negative).

  • send charts matching = * includes one or more space separated patterns, using * as wildcard (any number of times within each pattern). The patterns are checked against both chart id and chart name. A pattern starting with ! gives a negative match. So to match all charts named apps.* except charts ending in *reads, use !*reads apps.* (so, the order is important: the first pattern matching the chart id or the chart name will be used - positive or negative).

  • send names instead of ids = yes | no controls the metric names Netdata should send to the external database. Netdata supports names and IDs for charts and dimensions. Usually IDs are unique identifiers as read by the system and names are human friendly labels (also unique). Most charts and metrics have the same ID and name, but in several cases they are different: disks with device-mapper, interrupts, QoS classes, statsd synthetic charts, etc.

  • send configured labels = yes | no controls if labels defined in the [host labels] section in netdata.conf should be sent to the external database

  • send automatic labels = yes | no controls if automatially created labels, like _os_name or _architecture should be sent to the external database

Starting from Netdata v1.20 the host tags (defined in the [backend] section of netdata.conf) are parsed in accordance with a configured backend type and stored as host labels so that they can be reused in API responses and exporting connectors. The parsing is supported for graphite, json, opentsdb, and prometheus (default) backend types. You can check how the host tags were parsed using the /api/v1/info API call. But, keep in mind that backends subsystem is deprecated and will be deleted soon. Please move your existing tags to the [host labels] section.

Exporting engine monitoring

Netdata creates five charts in the dashboard, under the Netdata Monitoring section, to help you monitor the health and performance of the exporting engine itself:

  1. Buffered metrics, the number of metrics Netdata added to the buffer for dispatching them to the external database server.

  2. Exporting data size, the amount of data (in KB) Netdata added the buffer.

  3. Exporting operations, the number of operations performed by Netdata.

  4. Exporting thread CPU usage, the CPU resources consumed by the Netdata thread, that is responsible for sending the metrics to the external database server.

image

Exporting engine alarms

Netdata adds 3 alarms:

  1. exporting_last_buffering, number of seconds since the last successful buffering of exported data
  2. exporting_metrics_sent, percentage of metrics sent to the external database server
  3. exporting_metrics_lost, number of metrics lost due to repeating failures to contact the external database server

image

Last updated on