StatsD
The indicator data collected by the DDTrace agent will be sent to port 8125 of the DK through the StatsD data type.
This includes the JVM CPU, memory, threads, and class loading information of the JVM runtime, as well as various collected JMX indicators such as Kafka, Tomcat, RabbitMQ, etc.
Config¶
Preconditions¶
When DDTrace runs as an agent, there is no need for the user to specifically open the jmx port. If no port is opened, the agent will randomly open a local port.
DDTrace will collect JVM information by default. By default, it will be sent to 'localhost: 8125'
if k8s:
You can use dd.jmxfetch.<INTEGRATION_NAME>.enabled=true Enable the specified collector.
for INTEGRATION_NAME, You can check the default supported third-party software before.
For example, Tomcat or Kafka:
Collector Configuration¶
Go to the conf.d/statsd directory under the DataKit installation directory, copy statsd.conf.sample and name it statsd.conf. Examples are as follows:
[[inputs.statsd]]
## Collector alias.
# source = "statsd/-/-"
## Collect interval, default is 10 seconds. (optional)
# interval = '10s'
protocol = "udp"
## Address and port to host UDP listener on
service_address = ":8125"
## Tag request metric. Used for distinguish feed metric name.
## eg, DD_TAGS=source_key:tomcat,host_key:cn-shanghai-sq5ei
## eg, -Ddd.tags=source_key:tomcat,host_key:cn-shanghai-sq5ei
# statsd_source_key = "source_key"
# statsd_host_key = "host_key"
## Indicate whether report tag statsd_source_key and statsd_host_key.
# save_above_key = false
delete_gauges = true
delete_counters = true
delete_sets = true
delete_timings = true
## Counter metric is float in new Datakit version, set true if want be int.
# set_counter_int = false
## Percentiles to calculate for timing & histogram stats
percentiles = [50.0, 90.0, 99.0, 99.9, 99.95, 100.0]
## separator to use between elements of a statsd metric
metric_separator = "_"
## Parses tags in the datadog statsd format
## http://docs.datadoghq.com/guides/dogstatsd/
parse_data_dog_tags = true
## Parses datadog extensions to the statsd format
datadog_extensions = true
## Parses distributions metric as specified in the datadog statsd format
## https://docs.datadoghq.com/developers/metrics/types/?tab=distribution#definition
datadog_distributions = true
## We do not need following tags(they may create tremendous of time-series under influxdb's logic)
# Examples:
# "runtime-id", "metric-type"
drop_tags = [ ]
# All metric-name prefixed with 'jvm_' are set to influxdb's measurement 'jvm'
# All metric-name prefixed with 'stats_' are set to influxdb's measurement 'stats'
# Examples:
# "stats_:stats", "jvm_:jvm", "tomcat_:tomcat",
metric_mapping = [ ]
## Number of UDP messages allowed to queue up, once filled,
## the statsd server will start dropping packets, default is 128.
# allowed_pending_messages = 128
## Number of timing/histogram values to track per-measurement in the
## calculation of percentiles. Raising this limit increases the accuracy
## of percentiles but also increases the memory usage and cpu time.
percentile_limit = 1000
## Max duration (TTL) for each metric to stay cached/reported without being updated.
#max_ttl = "1000h"
[inputs.statsd.tags]
# some_tag = "some_value"
# more_tag = "some_other_value"
After configuration, restart DataKit.
The collector can now be turned on by configMap injection collector configuration.
Info
If find lot of Feed: io busy in the log, can configure interval='1s', minimum is 1s.
Tag data sources¶
If you want to tag the hosts captured by DDTrace, you can do so by injecting tags:
- You can use environment variables, i.e.
DD_TAGS, for example:DD_TAGS=source_key:tomcat,host_key:cn-shanghai-sq5ei - You can use the command line, i.e.
dd.tags, for example：-Ddd.tags=source_key:tomcat,host_key:cn-shanghai-sq5ei
In the above example, you need to specify that the key for source is source_key and the key for host is host_key in the Datakit configuration. You can change it to something else, but you have to make sure that the field names in Datakit are the same as the field names in DDTrace.
The end result is that you can see statsd/tomcat/cn-shanghai-sq5ei in the datakit monitor, which distinguishes it from other data sources reported to the statsd collector. If the above configuration is not done, then the default display is seen on the datakit monitor: statsd/-/-.
In addition, there is a configuration switch save_above_key that determine whether the tags corresponding to statsd_source_key and statsd_host_key are reported to the center. The default is not to report(false).
Metric¶
Statsd has no measurement definition at present, and all metrics are subject to the metrics sent by the network.
For example, if Tomcat or Kafka uses the default indicator set, GitHub can view all indicator sets