Configurationedit

Utilize configuration options to adapt the Elastic APM agent to your needs. One way to configure settings is with the php.ini file:

elastic_apm.server_url=http://localhost:8200
elastic_apm.service_name="My service"

The agent can also be configured using environment variables:

export ELASTIC_APM_SERVER_URL="http://localhost:8200"
export ELASTIC_APM_SERVICE_NAME="My service"

Elasticsearch Service on Elastic Cloudedit

The agent can be configured to send data to an Elasticsearch Service APM instance by setting the server_url and secret_token options to the corresponding values found in the APM & Fleet section of Elastic Cloud.

Configure the agent, for example via php.ini:

elastic_apm.server_url=APM_SERVER_URL
elastic_apm.secret_token=TOKEN
elastic_apm.service_name=SERVICE_NAME

Loggingedit

The easiest way to configure the logging is by using the log_level_syslog configuration option.

Available log levels are:

OFF
CRITICAL
ERROR
WARNING
INFO
DEBUG
TRACE

For example, if you specify a WARNING log level, only log records with levels WARNING, ERROR, and CRITICAL will be emitted.

OFF is only used to disable agent logging.

The agent supports logging to the following sinks: syslog and stderr. Control the level of logging for individual sinks with the log_level_syslog and log_level_stderr options. When a sink-specific logging level is not explicitly set, the fallback setting log_level will be used.

For example, the following configuration sets the log level to WARNING for all the sinks:

elastic_apm.log_level=WARNING

Alternatively, the following configuration sets log level to WARNING for all the sinks except for syslog, where the log level is set to TRACE.

elastic_apm.log_level=WARNING
elastic_apm.log_level_syslog=TRACE

Duration formatedit

The duration format is used for configuration options like timeouts. The units are provided as a suffix either directly after the number or separated by any amount of whitespace. The units are case insensitive so they can be provided in either lower, upper or even mixed case.

Supported units:

Example: 5m (interpreted as 5 minutes) or 10 mS (interpreted as 10 milliseconds).

Default units: Each configuration option with value specifiying a duration has default units. This allows omitting the units in option value. For example if option’s default units are s (seconds) then value 10 is interpreted as 10 seconds.

Size formatedit

The size format is used for options such as maximum buffer sizes. The units are provided as a suffix either directly after the number or separated by any amount of whitespace. The units are case insensitive so they can be provided in either lower, upper or even mixed case.

Supported units:

Example: 34KB (interpreted as 34 kilobytes) or 78 mB (interpreted as 78 megabytes).

Default units: Each configuration option with value specifiying a size has default units. This allows omitting the units in option value. For example if option’s default units are KB (kilobytes) then value 10 is interpreted as 10 kilobytes.

Wildcardedit

Some options (for example url_groups) support use of wildcard. A valid value for such configuration options is a comma separated list of wildcard expressions. Only the wildcard *, which matches zero or more characters, is supported.

Examples: *foo*, /foo/*/bar, /*/baz*.

Matching is case insensitive by default. Prepending an element with (?-i) makes the matching case sensitive. For example (?-i)/bar, /foo matches /bar` and /FOO but it doesn't match `/BAR on the other hand (?-i)/bar, (?-i)/foo matches /bar and /foo but doesn’t match neither /BAR nor /FOO.

Whitespace around commas separating wildcard expressions in the list is ignored. For example foo , bar is the same as foo,bar. On the other hand whitespace inside wildcard expressions is significant. For example *a b* matches a string only if it contains a followed by space and then b.

The input string is matched against wildcard expressions in the order they are listed and the first expression that matches is selected.

When configuration option is intended to matched against a input URL (for example url_groups and transaction_ignore_urls) only path part of the URL is tested against wildcard expressions. Other parts of the URL (such as query string, etc.) are not taken into account so including them in the wildcard expressions might lead to unexpected result. For example /user/* matches http://my_site.com/user/Alice?lang=en while /user/*?lang=* does not match http://my_site.com/user/Alice?lang=en