<?xml version="1.0" encoding="utf-8" ?>
<!DOCTYPE chapter SYSTEM "chapter.dtd">
<chapter>
<header>
<copyright>
<year>2017</year><year>2018</year>
<holder>Ericsson AB. All Rights Reserved.</holder>
</copyright>
<legalnotice>
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
</legalnotice>
<title>Logging</title>
<prepared></prepared>
<docno></docno>
<date></date>
<rev></rev>
<file>logger_chapter.xml</file>
</header>
<p>Erlang/OTP 21.0 provides a standard API for logging
through <c>Logger</c>, which is part of the Kernel
application. Logger consists of the API for issuing log events,
and a customizable backend where log handlers, filters and
formatters can be plugged in.</p>
<p>By default, the Kernel application installs one log handler at
system start. This handler is named <c>default</c>. It receives
and processes standard log events produced by the Erlang runtime
system, standard behaviours and different Erlang/OTP
applications. The log events are by default written to the
terminal.</p>
<p>You can also configure the system so that the default handler
prints log events to a single file, or to a set of wrap logs
via <seealso marker="disk_log"><c>disk_log</c></seealso>.</p>
<p>By configuration, you can also modify or disable the default
handler, replace it by a custom handler, and install additional
handlers.</p>
<note>
<p>Since Logger is new in Erlang/OTP 21.0, we do reserve the right
to introduce changes to the Logger API and functionality in
patches following this release. These changes might or might not
be backwards compatible with the initial version.</p>
</note>
<section>
<title>Overview</title>
<p>A <em>log event</em> consists of a <em>log level</em>, the
<em>message</em> to be logged, and <em>metadata</em>.</p>
<p>The Logger backend forwards log events from the API, first
through a set of <em>primary filters</em>, then through a set of
secondary filters attached to each log handler. The secondary
filters are in the following named <em>handler filters</em>.</p>
<p>Each filter set consists of a <em>log level check</em>,
followed by zero or more <em>filter functions</em>.</p>
<p>The following figure shows a conceptual overview of Logger. The
figure shows two log handlers, but any number of handlers can be
installed.</p>
<!-- The image is edited with dia in logger_arch.dia file,
and .png file generated with make target 'png'. -->
<image file="logger_arch.png">
<icaption>Conceptual Overview</icaption>
</image>
<p>Log levels are expressed as atoms. Internally in Logger, the
atoms are mapped to integer values, and a log event passes the
log level check if the integer value of its log level is less
than or equal to the currently configured log level. That is,
the check passes if the event is equally or more severe than the
configured level. See section <seealso marker="#log_level">Log
Level</seealso> for a listing and description of all log
levels.</p>
<p>The primary log level can be overridden by a log level
configured per module. This is to, for instance, allow more
verbose logging from a specific part of the system.</p>
<p>Filter functions can be used for more sophisticated filtering
than the log level check provides. A filter function can stop or
pass a log event, based on any of the event's contents. It can
also modify all parts of the log event. See see
section <seealso marker="#filters">Filters</seealso> for more
details.</p>
<p>If a log event passes through all primary filters and all
handler filters for a specific handler, Logger forwards the
event to the <em>handler callback</em>. The handler formats and
prints the event to its destination. See
section <seealso marker="#handlers">Handlers</seealso> for more
details.</p>
<p>Everything up to and including the call to the handler
callbacks is executed on the client process, that is, the
process where the log event was issued. It is up to the handler
implementation if other processes are involved or not.</p>
<p>The handlers are called in sequence, and the order is not
defined.</p>
</section>
<section>
<marker id="logger_api"/>
<title>Logger API</title>
<p>The API for logging consists of a set
of <seealso marker="logger#macros">macros</seealso>, and a set
of functions on the form <c>logger:Level/1,2,3</c>, which are
all shortcuts
for <seealso marker="logger#log-2">
<c>logger:log(Level,Arg1[,Arg2[,Arg3]])</c></seealso>.</p>
<p>The difference between using the macros and the exported
functions is that macros add location (originator) information
to the metadata, and performs lazy evaluation by wrapping the
logger call in a case statement, so it is only evaluated if the
log level of the event passes the primary log level check.</p>
<section>
<marker id="log_level"/>
<title>Log Level</title>
<p>The log level indicates the severity of a event. In
accordance with the Syslog protocol,
<url href="https://www.ietf.org/rfc/rfc5424.txt">RFC
5424</url>, eight log levels can be specified. The following
table lists all possible log levels by name (atom), integer
value, and description:</p>
<table align="left">
<row>
<cell><strong>Level</strong></cell>
<cell align="center"><strong>Integer</strong></cell>
<cell><strong>Description</strong></cell>
</row>
<row>
<cell>emergency</cell>
<cell align="center">0</cell>
<cell>system is unusable</cell>
</row>
<row>
<cell>alert</cell>
<cell align="center">1</cell>
<cell>action must be taken immediately</cell>
</row>
<row>
<cell>critical</cell>
<cell align="center">2</cell>
<cell>critical conditions</cell>
</row>
<row>
<cell>error</cell>
<cell align="center">3</cell>
<cell>error conditions</cell>
</row>
<row>
<cell>warning</cell>
<cell align="center">4</cell>
<cell>warning conditions</cell>
</row>
<row>
<cell>notice</cell>
<cell align="center">5</cell>
<cell>normal but significant conditions</cell>
</row>
<row>
<cell>info</cell>
<cell align="center">6</cell>
<cell>informational messages</cell>
</row>
<row>
<cell>debug</cell>
<cell align="center">7</cell>
<cell>debug-level messages</cell>
</row>
<tcaption>Log Levels</tcaption>
</table>
<p>Notice that the integer value is only used internally in
Logger. In the API, you must always use the atom. To compare
the severity of two log levels,
use <seealso marker="logger#compare_levels-2">
<c>logger:compare_levels/2</c></seealso>.</p>
</section>
<section>
<marker id="log_message"/>
<title>Log Message</title>
<p>The log message contains the information to be logged. The
message can consist of a format string and arguments (given as
two separate parameters in the Logger API), a string or a
report. The latter, which is either a map or a key-value list,
can be accompanied by a <em>report callback</em> specified in
the log event's <seealso marker="#metadata">metadata</seealso>.
The report callback is a convenience function that
the <seealso marker="#formatters">formatter</seealso> can use
to convert the report to a format string and arguments. The
formatter can also use its own conversion function, if no
callback is provided, or if a customized formatting is
desired.</p>
<p>Example, format string and arguments:</p>
<code>logger:error("The file does not exist: ~ts",[Filename])</code>
<p>Example, string:</p>
<code>logger:notice("Something strange happened!")</code>
<p>Example, report, and metadata with report callback:</p>
<code>
logger:debug(#{got => connection_request, id => Id, state => State},
#{report_cb => fun(R) -> {"~p",[R]} end})</code>
<p>The log message can also be provided through a fun for lazy
evaluation. The fun is only evaluated if the primary log level
check passes, and is therefore recommended if it is expensive
to generate the message. The lazy fun must return a string, a
report, or a tuple with format string and arguments.</p>
</section>
<section>
<title>Metadata</title>
<p>Metadata contains additional data associated with a log
message. Logger inserts some metadata fields by default, and
the client can add custom metadata in two different ways:</p>
<taglist>
<tag>Set process metadata</tag>
<item>
<p>Process metadata is set and updated
with <seealso marker="logger#set_process_metadata-1">
<c>logger:set_process_metadata/1</c></seealso>
and <seealso marker="logger#update_process_metadata-1">
<c>logger:update_process metadata/1</c></seealso>,
respectively. This metadata applies to the process on
which these calls are made, and Logger adds the metadata
to all log events issued on that process.</p>
</item>
<tag>Add metadata to a specific log event</tag>
<item>
<p>Metadata associated with one specific log event is given
as the last parameter to the log macro or Logger API
function when the event is issued. For example:</p>
<code>?LOG_ERROR("Connection closed",#{context => server})</code>
</item>
</taglist>
<p>See the description of
the <seealso marker="logger#type-metadata">
<c>logger:metadata()</c></seealso> type for information
about which default keys Logger inserts, and how the different
metadata maps are merged.</p>
</section>
</section>
<section>
<marker id="filter"/>
<title>Filters</title>
<p>Filters can be primary, or attached to a specific
handler. Logger calls the primary filters first, and if they all
pass, it calls the handler filters for each handler. Logger
calls the handler callback only if all filters attached to the
handler in question also pass.</p>
<p>A filter is defined as:</p>
<pre>{FilterFun, Extra}</pre>
<p>where <c>FilterFun</c> is a function of arity 2,
and <c>Extra</c> is any term. When applying the filter, Logger
calls the function with the log event as the first argument,
and the value of <c>Extra</c> as the second
argument. See <seealso marker="logger#type-filter">
<c>logger:filter()</c></seealso> for type definitions.</p>
<p>The filter function can return <c>stop</c>, <c>ignore</c> or
the (possibly modified) log event.</p>
<p>If <c>stop</c> is returned, the log event is immediately
discarded. If the filter is primary, no handler filters or
callbacks are called. If it is a handler filter, the
corresponding handler callback is not called, but the log event
is forwarded to filters attached to the next handler, if
any.</p>
<p>If the log event is returned, the next filter function is
called with the returned value as the first argument. That is,
if a filter function modifies the log event, the next filter
function receives the modified event. The value returned from
the last filter function is the value that the handler callback
receives.</p>
<p>If the filter function returns <c>ignore</c>, it means that it
did not recognize the log event, and thus leaves to other
filters to decide the event's destiny.</p>
<p>The configuration option <c>filter_default</c> specifies the
behaviour if all filter functions return <c>ignore</c>, or if no
filters exist. <c>filter_default</c> is by default set
to <c>log</c>, meaning that if all existing filters ignore a log
event, Logger forwards the event to the handler
callback. If <c>filter_default</c> is set to <c>stop</c>, Logger
discards such events.</p>
<p>Primary filters are added
with <seealso marker="logger#add_primary_filter-2">
<c>logger:add_primary_filter/2</c></seealso>
and removed
with <seealso marker="logger#remove_primary_filter-1">
<c>logger:remove_primary_filter/1</c></seealso>. They can also
be added at system start via the Kernel configuration
parameter <seealso marker="#logger_parameter"><c>logger</c></seealso>.</p>
<p>Handler filters are added
with <seealso marker="logger#add_handler_filter-3">
<c>logger:add_handler_filter/3</c></seealso>
and removed
with <seealso marker="logger#remove_handler_filter-2">
<c>logger:remove_handler_filter/2</c></seealso>. They can also
be specified directly in the configuration when adding a handler
with <seealso marker="logger#add_handler/3">
<c>logger:add_handler/3</c></seealso>
or via the Kernel configuration
parameter <seealso marker="#logger_parameter"><c>logger</c></seealso>.</p>
<p>To see which filters are currently installed in the system,
use <seealso marker="logger#get_config-0">
<c>logger:get_config/0</c></seealso>,
or <seealso marker="logger#get_primary_config-0">
<c>logger:get_primary_config/0</c></seealso>
and <seealso marker="logger#get_handler_config-1">
<c>logger:get_handler_config/1</c></seealso>. Filters are
listed in the order they are applied, that is, the first
filter in the list is applied first, and so on.</p>
<p>For convenience, the following built-in filters exist:</p>
<taglist>
<tag><seealso marker="logger_filters#domain-2">
<c>logger_filters:domain/2</c></seealso></tag>
<item>
<p>Provides a way of filtering log events based on a
<c>domain</c> field in <c>Metadata</c>.</p>
</item>
<tag><seealso marker="logger_filters#level-2">
<c>logger_filters:level/2</c></seealso></tag>
<item>
<p>Provides a way of filtering log events based on the log
level.</p>
</item>
<tag><seealso marker="logger_filters#progress-2">
<c>logger_filters:progress/2</c></seealso></tag>
<item>
<p>Stops or allows progress reports from <c>supervisor</c>
and <c>application_controller</c>.</p>
</item>
<tag><seealso marker="logger_filters#remote_gl-2">
<c>logger_filters:remote_gl/2</c></seealso></tag>
<item>
<p>Stops or allows log events originating from a process
that has its group leader on a remote node.</p>
</item>
</taglist>
</section>
<section>
<marker id="handlers"/>
<title>Handlers</title>
<p>A handler is defined as a module exporting at least the
following callback function:</p>
<pre><seealso marker="logger#HModule:log-2">log(LogEvent, Config) -> void()</seealso></pre>
<p>This function is called when a log event has passed through all
primary filters, and all handler filters attached to the handler
in question. The function call is executed on the client
process, and it is up to the handler implementation if other
processes are involved or not.</p>
<p>Logger allows adding multiple instances of a handler
callback. That is, if a callback module implementation allows
it, you can add multiple handler instances using the same
callback module. The different instances are identified by
unique handler identities.</p>
<p>In addition to the mandatory callback function <c>log/2</c>, a
handler module can export the optional callback
functions <c>adding_handler/1</c>, <c>changing_config/2</c>
and <c>removing_handler/1</c>. See
section <seealso marker="logger#handler_callback_functions">Handler
Callback Functions</seealso> in the logger(3) manual page for
more information about these function.</p>
<p>The following built-in handlers exist:</p>
<taglist>
<tag><c>logger_std_h</c></tag>
<item>
<p>This is the default handler used by OTP. Multiple instances
can be started, and each instance will write log events to a
given destination, terminal or file.</p>
</item>
<tag><c>logger_disk_log_h</c></tag>
<item>
<p>This handler behaves much like <c>logger_std_h</c>, except it uses
<seealso marker="disk_log"><c>disk_log</c></seealso> as its
destination.</p>
</item>
<tag><marker id="ErrorLoggerManager"/><c>error_logger</c></tag>
<item>
<p>This handler is provided for backwards compatibility
only. It is not started by default, but will be
automatically started the first time an <c>error_logger</c>
event handler is added
with <seealso marker="error_logger#add_report_handler-1">
<c>error_logger:add_report_handler/1,2</c></seealso>.</p>
<p>The old <c>error_logger</c> event handlers in STDLIB and
SASL still exist, but they are not added by Erlang/OTP 21.0
or later.</p>
</item>
</taglist>
</section>
<section>
<marker id="formatters"/>
<title>Formatters</title>
<p>A formatter can be used by the handler implementation to do the
final formatting of a log event, before printing to the
handler's destination. The handler callback receives the
formatter information as part of the handler configuration,
which is passed as the second argument
to <seealso marker="logger#HModule:log-2">
<c>HModule:log/2</c></seealso>.</p>
<p>The formatter information consist of a formatter
module, <c>FModule</c> and its
configuration, <c>FConfig</c>. <c>FModule</c> must export the
following function, which can be called by the handler:</p>
<pre><seealso marker="logger#FModule:format-2">format(LogEvent,FConfig)
-> FormattedLogEntry</seealso></pre>
<p>The formatter information for a handler is set as a part of its
configuration when the handler is added. It can also be changed
during runtime
with <seealso marker="logger#set_handler_config-3">
<c>logger:set_handler_config(HandlerId,formatter,{FModule,FConfig})</c>
</seealso>, which overwrites the current formatter information,
or with <seealso marker="logger#update_formatter_config-2">
<c>logger:update_formatter_config/2,3</c></seealso>, which
only modifies the formatter configuration.</p>
<p>If the formatter module exports the optional callback
function <seealso marker="logger#FModule:check_config-1">
<c>check_config(FConfig)</c></seealso>, Logger calls this
function when the formatter information is set or modified, to
verify the validity of the formatter configuration.</p>
<p>If no formatter information is specified for a handler, Logger
uses <c>logger_formatter</c> as default. See
the <seealso marker="logger_formatter"><c>logger_formatter(3)</c></seealso>
manual page for more information about this module.</p>
</section>
<section>
<title>Configuration</title>
<p>At system start, Logger is configured through Kernel
configuration parameters. The parameters that apply to Logger
are described in
section <seealso marker="#kernel_config_params">Kernel
Configuration Parameters</seealso>. Examples are found in
section <seealso marker="#config_examples">Configuration
Examples</seealso>.</p>
<p>During runtime, Logger configuration is changed via API
functions. See
section <seealso marker="logger#configuration_API">Configuration
API Functions</seealso> in the <c>logger(3)</c> manual page.</p>
<section>
<title>Primary Logger Configuration</title>
<p>Logger API functions that apply to the primary Logger
configuration are:</p>
<list>
<item><seealso marker="logger#get_primary_config-0">
<c>get_primary_config/0</c></seealso></item>
<item><seealso marker="logger#set_primary_config-1">
<c>set_primary_config/1,2</c></seealso></item>
<item><seealso marker="logger#update_primary_config-1">
<c>update_primary_config/1</c></seealso></item>
<item><seealso marker="logger#add_primary_filter-2">
<c>add_primary_filter/2</c></seealso></item>
<item><seealso marker="logger#remove_primary_filter-1">
<c>remove_primary_filter/1</c></seealso></item>
</list>
<p>The primary Logger configuration is a map with the following
keys:</p>
<taglist>
<tag><marker id="primary_level"/>
<c>level = </c><seealso marker="logger#type-level">
<c>logger:level()</c></seealso><c> | all | none</c></tag>
<item>
<p>Specifies the primary log level, that is, log event that
are equally or more severe than this level, are forwarded
to the primary filters. Less severe log events are
immediately discarded.</p>
<p>See section <seealso marker="#log_level">Log
Level</seealso> for a listing and description of
possible log levels.</p>
<p>The initial value of this option is set by the Kernel
configuration parameter <seealso marker="#logger_level">
<c>logger_level</c></seealso>. It is changed during
runtime with <seealso marker="logger#set_primary_config-2">
<c>logger:set_primary_config(level,Level)</c></seealso>.</p>
<p>Defaults to <c>info</c>.</p>
</item>
<tag><c>filters = [{FilterId,Filter}]</c></tag>
<item>
<p>Specifies the primary filters.</p>
<list>
<item><c>FilterId = </c><seealso marker="logger#type-filter_id">
<c>logger:filter_id()</c></seealso></item>
<item><c>Filter = </c><seealso marker="logger#type-filter">
<c>logger:filter()</c></seealso></item>
</list>
<p>The initial value of this option is set by the Kernel
configuration
parameter <seealso marker="#logger_parameter"><c>logger</c></seealso>.
During runtime, primary filters are added and removed with
<seealso marker="logger#add_primary_filter-2">
<c>logger:add_primary_filter/2</c></seealso> and
<seealso marker="logger#remove_primary_filter-1">
<c>logger:remove_primary_filter/1</c></seealso>,
respectively.</p>
<p>See section <seealso marker="#filters">Filters</seealso>
for more detailed information.</p>
<p>Defaults to <c>[]</c>.</p>
</item>
<tag><c>filter_default = log | stop</c></tag>
<item>
<p>Specifies what happens to a log event if all filters
return <c>ignore</c>, or if no filters exist.</p>
<p>See section <seealso marker="#filters">Filters</seealso>
for more information about how this option is used.</p>
<p>Defaults to <c>log</c>.</p>
</item>
</taglist>
</section>
<section>
<marker id="handler_configuration"/>
<title>Handler Configuration</title>
<p>Logger API functions that apply to handler configuration
are:</p>
<list>
<item><seealso marker="logger#get_handler_config-0">
<c>get_handler_config/0,1</c></seealso></item>
<item><seealso marker="logger#set_handler_config-2">
<c>set_handler_config/2,3</c></seealso></item>
<item><seealso marker="logger#update_handler_config-2">
<c>update_handler_config/2</c></seealso></item>
<item><seealso marker="logger#add_handler_filter-3">
<c>add_handler_filter/3</c></seealso></item>
<item><seealso marker="logger#remove_handler_filter-2">
<c>remove_handler_filter/2</c></seealso></item>
<item><seealso marker="logger#update_formatter_config-2">
<c>update_formatter_config/2,3</c></seealso></item>
</list>
<p>The configuration for a handler is a map with the following keys:</p>
<taglist>
<tag><c>id = </c><seealso marker="logger#type-handler_id">
<c>logger:handler_id()</c></seealso></tag>
<item>
<p>Automatically inserted by Logger. The value is the same
as the <c>HandlerId</c> specified when adding the handler,
and it cannot be changed.</p>
</item>
<tag><c>module = module()</c></tag>
<item>
<p>Automatically inserted by Logger. The value is the same
as the <c>Module</c> specified when adding the handler,
and it cannot be changed.</p>
</item>
<tag><c>level = </c><seealso marker="logger#type-level">
<c>logger:level()</c></seealso><c> | all | none</c></tag>
<item>
<p>Specifies the log level for the handler, that is, log
events that are equally or more severe than this level,
are forwarded to the handler filters for this
handler.</p>
<p>See section <seealso marker="#log_level">Log
Level</seealso> for a listing and description of
possible log levels.</p>
<p>The log level is specified when adding the handler, or
changed during runtime with, for
instance, <seealso marker="logger#set_handler_config/3">
<c>logger:set_handler_config(HandlerId,level,Level)</c></seealso>.
</p>
<p>Defaults to <c>all</c>.</p>
</item>
<tag><c>filters = [{FilterId,Filter}]</c></tag>
<item>
<p>Specifies the handler filters.</p>
<list>
<item><c>FilterId = </c><seealso marker="logger#type-filter_id">
<c>logger:filter_id()</c></seealso></item>
<item><c>Filter = </c><seealso marker="logger#type-filter">
<c>logger:filter()</c></seealso></item>
</list>
<p>Handler filters are specified when adding the handler,
or added or removed during runtime with
<seealso marker="logger#add_handler_filter-3">
<c>logger:add_handler_filter/3</c></seealso> and
<seealso marker="logger#remove_handler_filter-2">
<c>logger:remove_handler_filter/2</c></seealso>,
respectively.</p>
<p>See <seealso marker="#filters">Filters</seealso> for more
detailed information.</p>
<p>Defaults to <c>[]</c>.</p>
</item>
<tag><c>filter_default = log | stop</c></tag>
<item>
<p>Specifies what happens to a log event if all filters
return <c>ignore</c>, or if no filters exist.</p>
<p>See section <seealso marker="#filters">Filters</seealso>
for more information about how this option is used.</p>
<p>Defaults to <c>log</c>.</p>
</item>
<tag><c>formatter = {FormatterModule,FormatterConfig}</c></tag>
<item>
<p>Specifies a formatter that the handler can use for
converting the log event term to a printable string.</p>
<list>
<item><c>FormatterModule = module()</c></item>
<item><c>FormatterConfig = </c>
<seealso marker="logger#type-formatter_config">
<c>logger:formatter_config()</c></seealso></item>
</list>
<p>The formatter information is specified when adding the
handler. The formatter configuration can be changed during
runtime
with <seealso marker="logger#update_formatter_config-2">
<c>logger:update_formatter_config/2,3</c></seealso>,
or the complete formatter information can be overwritten
with, for
instance, <seealso marker="logger#set_handler_config-3">
<c>logger:set_handler_config/3</c></seealso>.</p>
<p>See
section <seealso marker="#formatters">Formatters</seealso>
for more detailed information.</p>
<p>Defaults
to <c>{logger_formatter,DefaultFormatterConfig}</c>. See
the <seealso marker="logger_formatter">
<c>logger_formatter(3)</c></seealso> manual page for
information about this formatter and its default
configuration.</p>
</item>
<tag><c>config = term()</c></tag>
<item>
<p>Handler specific configuration, that is, configuration
data related to a specific handler implementation.</p>
<p>The configuration for the built-in handlers is described
in
the <seealso marker="logger_std_h"><c>logger_std_h(3)</c></seealso>
and
<seealso marker="logger_disk_log_h"><c>logger_disk_log_h(3)</c>
</seealso> manual pages.</p>
</item>
</taglist>
<p>Notice that <c>level</c> and <c>filters</c> are obeyed by
Logger itself before forwarding the log events to each
handler, while <c>formatter</c> and all handler specific
options are left to the handler implementation.</p>
</section>
<section>
<marker id="kernel_config_params"/>
<title>Kernel Configuration Parameters</title>
<p>The following Kernel configuration parameters apply to
Logger:</p>
<taglist>
<tag><marker id="logger_parameter"/><c>logger = [Config]</c></tag>
<item>
<p>Specifies the configuration
for <seealso marker="logger">Logger</seealso>, except the
primary log level, which is specified
with <seealso marker="#logger_level"><c>logger_level</c></seealso>,
and the compatibility
with <seealso marker="sasl:error_logging">SASL Error
Logging</seealso>, which is specified
with <seealso marker="#logger_sasl_compatible">
<c>logger_sasl_compatible</c></seealso>.</p>
<p>With this parameter, you can modify or disable the default
handler, add custom handlers and primary logger filters, and
set log levels per module.</p>
<p><c>Config</c> is any (zero or more) of the following:</p>
<taglist>
<tag><c>{handler, default, undefined}</c></tag>
<item>
<p>Disables the default handler. This allows another
application to add its own default handler.</p>
<p>Only one entry of this type is allowed.</p>
</item>
<tag><c>{handler, HandlerId, Module, HandlerConfig}</c></tag>
<item>
<p>If <c>HandlerId</c> is <c>default</c>, then this entry
modifies the default handler, equivalent to calling</p>
<pre><seealso marker="logger#set_handler_config-2">
logger:set_handler_config(default, Module, HandlerConfig)
</seealso></pre>
<p>For all other values of <c>HandlerId</c>, this entry
adds a new handler, equivalent to calling</p>
<pre><seealso marker="logger:add_handler/3">
logger:add_handler(HandlerId, Module, HandlerConfig)
</seealso></pre>
<p>Multiple entries of this type are allowed.</p></item>
<tag><c>{filters, FilterDefault, [Filter]}</c></tag>
<item>
<p>Adds the specified primary filters.</p>
<list>
<item><c>FilterDefault = log | stop</c></item>
<item><c>Filter = {FilterId, {FilterFun, FilterConfig}}</c></item>
</list>
<p>Equivalent to calling</p>
<pre><seealso marker="logger#add_primary_filter/2">
logger:add_primary_filter(FilterId, {FilterFun, FilterConfig})
</seealso></pre>
<p>for each <c>Filter</c>.</p>
<p><c>FilterDefault</c> specifies the behaviour if all
primary filters return <c>ignore</c>, see
section <seealso marker="#filters">Filters</seealso>.</p>
<p>Only one entry of this type is allowed.</p>
</item>
<tag><c>{module_level, Level, [Module]}</c></tag>
<item>
<p>Sets module log level for the given modules. Equivalent
to calling</p>
<pre><seealso marker="logger#set_module_level/2">
logger:set_module_level(Module, Level)</seealso></pre>
<p>for each <c>Module</c>.</p>
<p>Multiple entries of this type are allowed.</p>
</item>
</taglist>
<p>See
section <seealso marker="#config_examples">Configuration
Examples</seealso> for examples using the <c>logger</c>
parameter for system configuration.</p>
</item>
<tag><marker id="logger_level"/>
<c>logger_level = Level</c></tag>
<item>
<p>Specifies the primary log level. See
the <seealso marker="kernel_app#logger_level"><c>kernel(6)</c></seealso>
manual page for more information about this parameter.</p>
</item>
<tag><marker id="logger_sasl_compatible"/>
<c>logger_sasl_compatible = true | false</c></tag>
<item>
<p>Specifies Logger's compatibility
with <seealso marker="sasl:error_logging">SASL Error
Logging</seealso>. See
the <seealso marker="kernel_app#logger_sasl_compatible">
<c>kernel(6)</c></seealso> manual page for more
information about this parameter.</p>
</item>
</taglist>
</section>
<section>
<marker id="config_examples"/>
<title>Configuration Examples</title>
<p>The value of the Kernel configuration parameter <c>logger</c>
is a list of tuples. It is possible to write the term on the
command line when starting an erlang node, but as the term
grows, a better approach is to use the system configuration
file. See
the <seealso marker="config"><c>config(4)</c></seealso> manual
page for more information about this file.</p>
<p>Each of the following examples shows a simple system
configuration file that configures Logger according to the
description.</p>
<p>Modify the default handler to print to a file instead of
<c>standard_io</c>:</p>
<code>
[{kernel,
[{logger,
[{handler, default, logger_std_h, % {handler, HandlerId, Module,
#{config => #{type => {file,"log/erlang.log"}}}} % Config}
]}]}].
</code>
<p>Modify the default handler to print each log event as a
single line:</p>
<code>
[{kernel,
[{logger,
[{handler, default, logger_std_h,
#{formatter => {logger_formatter, #{single_line => true}}}}
]}]}].
</code>
<p>Modify the default handler to print the pid of the logging
process for each log event:</p>
<code>
[{kernel,
[{logger,
[{handler, default, logger_std_h,
#{formatter => {logger_formatter,
#{template => [time," ",pid," ",msg,"\n"]}}}}
]}]}].
</code>
<p>Modify the default handler to only print errors and more
severe log events to "log/erlang.log", and add another handler
to print all log events to "log/debug.log".</p>
<code>
[{kernel,
[{logger,
[{handler, default, logger_std_h,
#{level => error,
config => #{type => {file, "log/erlang.log"}}}},
{handler, info, logger_std_h,
#{level => debug,
config => #{type => {file, "log/debug.log"}}}}
]}]}].
</code>
</section>
</section>
<section>
<marker id="compatibility"/>
<title>Backwards Compatibility with error_logger</title>
<p>Logger provides backwards compatibility with
<c>error_logger</c> in the following ways:</p>
<taglist>
<tag>API for Logging</tag>
<item>
<p>The <c>error_logger</c> API still exists, but should only
be used by legacy code. It will be removed in a later
release.</p>
<p>Calls
to <seealso marker="error_logger#error_report-1">
<c>error_logger:error_report/1,2</c></seealso>,
<seealso marker="error_logger#error_msg-1">
<c>error_logger:error_msg/1,2</c></seealso>, and
corresponding functions for warning and info messages, are
all forwarded to Logger as calls
to <seealso marker="logger#log-3">
<c>logger:log(Level,Report,Metadata)</c></seealso>.</p>
<p><c>Level = error | warning | info</c> and is taken
from the function name. <c>Report</c> contains the actual
log message, and <c>Metadata</c> contains additional
information which can be used for creating backwards
compatible events for legacy <c>error_logger</c> event
handlers, see
section <seealso marker="#legacy_event_handlers">Legacy
Event Handlers</seealso>.</p>
</item>
<tag>Output Format</tag>
<item>
<p>To get log events on the same format as produced
by <c>error_logger_tty_h</c> and <c>error_logger_file_h</c>,
use the default formatter, <c>logger_formatter</c>, with
configuration parameter <c>legacy_header</c> set
to <c>true</c>. This is the default configuration of
the <c>default</c> handler started by Kernel.</p>
</item>
<tag>Default Format of Log Events from OTP</tag>
<item>
<p>By default, all log events originating from within OTP,
except the former so called "SASL reports", look the same as
before.</p>
</item>
<tag><marker id="sasl_reports"/>SASL Reports</tag>
<item>
<p>By SASL reports we mean supervisor reports, crash reports
and progress reports.</p>
<p>Prior to Erlang/OTP 21.0, these reports were only logged
when the SASL application was running, and they were printed
trough SASL's own event handlers <c>sasl_report_tty_h</c>
and <c>sasl_report_file_h</c>.</p>
<p>The destination of these log events was configured by
<seealso marker="sasl:sasl_app#deprecated_error_logger_config">SASL
configuration parameters</seealso>.</p>
<p>Due to the specific event handlers, the output format
slightly differed from other log events.</p>
<p>As of Erlang/OTP 21.0, the concept of SASL reports is
removed, meaning that the default behaviour is as
follows:</p>
<list>
<item>Supervisor reports, crash reports, and progress reports
are no longer connected to the SASL application.</item>
<item>Supervisor reports and crash reports are issued
as <c>error</c> level log events, and are logged through
the default handler started by Kernel.</item>
<item>Progress reports are issued as <c>info</c> level log
events, and since the default primary log level
is <c>notice</c>, these are not logged by default. To
enable printing of progress reports, set
the <seealso marker="#primary_level">primary log
level</seealso> to <c>info</c>.</item>
<item>The output format is the same for all log
events.</item>
</list>
<p>If the old behaviour is preferred, the Kernel configuration
parameter <seealso marker="kernel_app#logger_sasl_compatible">
<c>logger_sasl_compatible</c></seealso> can be set
to <c>true</c>. The
<seealso marker="sasl:sasl_app#deprecated_error_logger_config">SASL
configuration parameters</seealso> can then be used as
before, and the SASL reports will only be printed if the
SASL application is running, through a second log handler
named <c>sasl</c>.</p>
<p>All SASL reports have a metadata field <c>domain</c> which
is set to <c>[otp,sasl]</c>. This field can be
used by filters to stop or allow the log events.</p>
<p>See section <seealso marker="sasl:error_logging">SASL User's
Guide</seealso> for more information about the old SASL
error logging functionality.</p>
</item>
<tag><marker id="legacy_event_handlers"/>Legacy Event Handlers</tag>
<item>
<p>To use event handlers written for <c>error_logger</c>, just
add your event handler with</p>
<code>
error_logger:add_report_handler/1,2.
</code>
<p>This automatically starts the error logger event manager,
and adds <c>error_logger</c> as a handler to Logger, with
the following configuration:</p>
<code>
#{level => info,
filter_default => log,
filters => []}.
</code>
<note>
<p>This handler ignores events that do not originate from
the <c>error_logger</c> API, or from within OTP. This
means that if your code uses the Logger API for logging,
then your log events will be discarded by this
handler.</p>
<p>The handler is not overload protected.</p>
</note>
</item>
</taglist>
</section>
<section>
<title>Error Handling</title>
<p>Logger does, to a certain extent, check its input data before
forwarding a log event to filters and handlers. It does,
however, not evaluate report callbacks, or check the validity of
format strings and arguments. This means that all filters and
handlers must be careful when formatting the data of a log
event, making sure that it does not crash due to bad input data
or faulty callbacks.</p>
<p>If a filter or handler still crashes, Logger will remove the
filter or handler in question from the configuration, and print
a short error message to the terminal. A debug event containing
the crash reason and other details is also issued.</p>
<p>See section <seealso marker="#log_message">Log
Message</seealso> for more information about report callbacks
and valid forms of log messages.</p>
</section>
<section>
<title>Example: Add a handler to log info events to file</title>
<p>When starting an Erlang node, the default behaviour is that all
log events on level <c>notice</c> or more severe, are logged to
the terminal via the default handler. To also log info events,
you can either change the primary log level to <c>info</c>:</p>
<pre>
1> <input>logger:set_primary_config(level, info).</input>
ok</pre>
<p>or set the level for one or a few modules only:</p>
<pre>
2> <input>logger:set_module_level(mymodule, info).</input>
ok</pre>
<p>This allows info events to pass through to the default handler,
and be printed to the terminal as well. If there are many info
events, it can be useful to print these to a file instead.</p>
<p>First, set the log level of the default handler
to <c>notice</c>, preventing it from printing info events to the
terminal:</p>
<pre>
3> <input>logger:set_handler_config(default, level, notice).</input>
ok</pre>
<p>Then, add a new handler which prints to file. You can use the
handler
module <seealso marker="logger_std_h"><c>logger_std_h</c></seealso>,
and specify type <c>{file,File}</c>.:</p>
<pre>
4> <input>Config = #{config => #{type => {file,"./info.log"}}, level => info}.</input>
#{config => #{type => {file,"./info.log"}},level => info}
5> <input>logger:add_handler(myhandler, logger_std_h, Config).</input>
ok</pre>
<p>Since <c>filter_default</c> defaults to <c>log</c>, this
handler now receives all log events. If you want info events
only in the file, you must add a filter to stop all non-info
events. The built-in
filter <seealso marker="logger_filters#level-2">
<c>logger_filters:level/2</c></seealso>
can do this:</p>
<pre>
6> <input>logger:add_handler_filter(myhandler, stop_non_info,
{fun logger_filters:level/2, {stop, neq, info}}).</input>
ok</pre>
<p>See section <seealso marker="#filters">Filters</seealso> for
more information about the filters and the <c>filter_default</c>
configuration parameter.</p>
</section>
<section>
<title>Example: Implement a handler</title>
<p>Section <seealso marker="logger#handler_callback_functions">Handler
Callback Functions</seealso> in the logger(3) manual page
describes the callback functions that can be implemented for a
Logger handler.</p>
<p>A handler callback module must export:</p>
<list>
<item><c>log(Log, Config)</c></item>
</list>
<p>It can optionally also export some, or all, of the following:</p>
<list>
<item><c>adding_handler(Config)</c></item>
<item><c>removing_handler(Config)</c></item>
<item><c>changing_config(OldConfig, NewConfig)</c></item>
</list>
<p>When a handler is added, by for example a call
to <seealso marker="logger#add_handler-3">
<c>logger:add_handler(Id, HModule, Config)</c></seealso>,
Logger first calls <c>HModule:adding_handler(Config)</c>. If
this function returns <c>{ok,Config1}</c>, Logger
writes <c>Config1</c> to the configuration database, and
the <c>logger:add_handler/3</c> call returns. After this, the
handler is installed and must be ready to receive log events as
calls to <c>HModule:log/2</c>.</p>
<p>A handler can be removed by calling
<seealso marker="logger#remove_handler-1">
<c>logger:remove_handler(Id)</c></seealso>. Logger calls
<c>HModule:removing_handler(Config)</c>, and removes the
handler's configuration from the configuration database.</p>
<p>When <seealso marker="logger#set_handler_config-2">
<c>logger:set_handler_config/2,3</c></seealso>
or <seealso marker="logger#update_handler_config/2">
<c>logger:update_handler_config/2</c></seealso> is called,
Logger
calls <c>HModule:changing_config(OldConfig, NewConfig)</c>. If
this function returns <c>{ok,NewConfig1}</c>, Logger
writes <c>NewConfig1</c> to the configuration database.</p>
<p>A simple handler that prints to the terminal can be implemented
as follows:</p>
<code>
-module(myhandler1).
-export([log/2]).
log(LogEvent, #{formatter := {FModule, FConfig}}) ->
io:put_chars(FModule:format(LogEvent, FConfig)).
</code>
<p>Notice that the above handler does not have any overload
protection, and all log events are printed directly from the
client process.</p>
<p>For information and examples of overload protection, please
refer to
section <seealso marker="#overload_protection">Protecting the
Handler from Overload</seealso>, and the implementation
of <seealso marker="logger_std_h"><c>logger_std_h</c></seealso>
and <seealso marker="logger_disk_log_h"><c>logger_disk_log_h</c>
</seealso>.</p>
<p>The following is a simpler example of a handler which logs to a
file through one single process:</p>
<code>
-module(myhandler2).
-export([adding_handler/1, removing_handler/1, log/2]).
-export([init/1, handle_call/3, handle_cast/2, terminate/2]).
adding_handler(Config) ->
MyConfig = maps:get(config,Config,#{file => "myhandler2.log"}),
{ok, Pid} = gen_server:start(?MODULE, MyConfig, []),
{ok, Config#{config => MyConfig#{pid => Pid}}}.
removing_handler(#{config := #{pid := Pid}}) ->
gen_server:stop(Pid).
log(LogEvent,#{config := #{pid := Pid}} = Config) ->
gen_server:cast(Pid, {log, LogEvent, Config}).
init(#{file := File}) ->
{ok, Fd} = file:open(File, [append, {encoding, utf8}]),
{ok, #{file => File, fd => Fd}}.
handle_call(_, _, State) ->
{reply, {error, bad_request}, State}.
handle_cast({log, LogEvent, Config}, #{fd := Fd} = State) ->
do_log(Fd, LogEvent, Config),
{noreply, State}.
terminate(_Reason, #{fd := Fd}) ->
_ = file:close(Fd),
ok.
do_log(Fd, LogEvent, #{formatter := {FModule, FConfig}}) ->
String = FModule:format(LogEvent, FConfig),
io:put_chars(Fd, String).
</code>
</section>
<section>
<marker id="overload_protection"/>
<title>Protecting the Handler from Overload</title>
<p>The default handlers, <seealso marker="logger_std_h">
<c>logger_std_h</c></seealso> and <seealso marker="logger_disk_log_h">
<c>logger_disk_log_h</c></seealso>, feature an overload protection
mechanism, which makes it possible for the handlers to survive,
and stay responsive, during periods of high load (when huge
numbers of incoming log requests must be handled).
The mechanism works as follows:</p>
<section>
<title>Message Queue Length</title>
<p>The handler process keeps track of the length of its message
queue and takes some form of action when the current length exceeds a
configurable threshold. The purpose is to keep the handler in, or to
as quickly as possible get the handler into, a state where it can
keep up with the pace of incoming log events. The memory use of the
handler must never grow larger and larger, since that will eventually
cause the handler to crash. These three thresholds, with associated
actions, exist:</p>
<taglist>
<tag><c>sync_mode_qlen</c></tag>
<item>
<p>As long as the length of the message queue is lower than this
value, all log events are handled asynchronously. This means that
the client process sending the log event, by calling a log function
in the <seealso marker="logger_chapter#logger_api">Logger API</seealso>,
does not wait for a response from the handler but continues
executing immediately after the event is sent. It is not affected
by the time it takes the handler to print the event to the log
device. If the message queue grows larger than this value,
the handler starts handling log events synchronously instead,
meaning that the client process sending the event must wait for a
response. When the handler reduces the message queue to a
level below the <c>sync_mode_qlen</c> threshold, asynchronous
operation is resumed. The switch from asynchronous to synchronous
mode can slow down the logging tempo of one, or a few, busy senders,
but cannot protect the handler sufficiently in a situation of many
busy concurrent senders.</p>
<p>Defaults to <c>10</c> messages.</p>
</item>
<tag><c>drop_mode_qlen</c></tag>
<item>
<p>When the message queue grows larger than this threshold, the
handler switches to a mode in which it drops all new events that
senders want to log. Dropping an event in this mode means that the
call to the log function never results in a message being sent to
the handler, but the function returns without taking any action.
The handler keeps logging the events that are already in its message
queue, and when the length of the message queue is reduced to a level
below the threshold, synchronous or asynchronous mode is resumed.
Notice that when the handler activates or deactivates drop mode,
information about it is printed in the log.</p>
<p>Defaults to <c>200</c> messages.</p>
</item>
<tag><c>flush_qlen</c></tag>
<item>
<p>If the length of the message queue grows larger than this threshold,
a flush (delete) operation takes place. To flush events, the handler
discards the messages in the message queue by receiving them in a
loop without logging. Client processes waiting for a response from a
synchronous log request receive a reply from the handler indicating
that the request is dropped. The handler process increases its
priority during the flush loop to make sure that no new events
are received during the operation. Notice that after the flush operation
is performed, the handler prints information in the log about how many
events have been deleted.</p>
<p>Defaults to <c>1000</c> messages.</p>
</item>
</taglist>
<p>For the overload protection algorithm to work properly, it is
required that:</p>
<p><c>sync_mode_qlen =< drop_mode_qlen =< flush_qlen</c></p>
<p>and that:</p>
<p><c>drop_mode_qlen > 1</c></p>
<p>To disable certain modes, do the following:</p>
<list>
<item>If <c>sync_mode_qlen</c> is set to <c>0</c>, all log events are handled
synchronously. That is, asynchronous logging is disabled.</item>
<item>If <c>sync_mode_qlen</c> is set to the same value as
<c>drop_mode_qlen</c>, synchronous mode is disabled. That is, the handler
always runs in asynchronous mode, unless dropping or flushing is invoked.</item>
<item>If <c>drop_mode_qlen</c> is set to the same value as <c>flush_qlen</c>,
drop mode is disabled and can never occur.</item>
</list>
<p>During high load scenarios, the length of the handler message queue
rarely grows in a linear and predictable way. Instead, whenever the
handler process is scheduled in, it can have an almost arbitrary number
of messages waiting in the message queue. It is for this reason that the overload
protection mechanism is focused on acting quickly, and quite drastically,
such as immediately dropping or flushing messages, when a large queue length
is detected.</p>
<p>The values of the previously listed thresholds can be specified by the user.
This way, a handler can be configured to, for example, not drop or flush
messages unless the message queue length of the handler process grows extremely
large. Notice that large amounts of memory can be required for the node under such
circumstances. Another example of user configuration is when, for performance
reasons, the client processes must never be blocked by synchronous log requests.
It is possible, perhaps, that dropping or flushing events is still acceptable, since
it does not affect the performance of the client processes sending the log events.</p>
<p>A configuration example:</p>
<code type="none">
logger:add_handler(my_standard_h, logger_std_h,
#{config => #{type => {file,"./system_info.log"},
sync_mode_qlen => 100,
drop_mode_qlen => 1000,
flush_qlen => 2000}}).
</code>
</section>
<section>
<title>Controlling Bursts of Log Requests</title>
<p>Large bursts of log events - many events received by the handler
under a short period of time - can potentially cause problems, such as:</p>
<list>
<item>Log files grow very large, very quickly.</item>
<item>Circular logs wrap too quickly so that important data is overwritten.</item>
<item>Write buffers grow large, which slows down file sync operations.</item>
</list>
<p>For this reason, both built-in handlers offer the possibility to specify the
maximum number of events to be handled within a certain time frame.
With this burst control feature enabled, the handler can avoid choking the log with
massive amounts of printouts. The configuration parameters are:</p>
<taglist>
<tag><c>burst_limit_enable</c></tag>
<item>
<p>Value <c>true</c> enables burst control and <c>false</c> disables it.</p>
<p>Defaults to <c>true</c>.</p>
</item>
<tag><c>burst_limit_max_count</c></tag>
<item>
<p>This is the maximum number of events to handle within a
<c>burst_limit_window_time</c> time frame. After the limit is
reached, successive events are dropped until the end of the time frame.</p>
<p>Defaults to <c>500</c> events.</p>
</item>
<tag><c>burst_limit_window_time</c></tag>
<item>
<p>See the previous description of <c>burst_limit_max_count</c>.</p>
<p>Defaults to <c>1000</c> milliseconds.</p>
</item>
</taglist>
<p>A configuration example:</p>
<code type="none">
logger:add_handler(my_disk_log_h, logger_disk_log_h,
#{config => #{file => "./my_disk_log",
burst_limit_enable => true,
burst_limit_max_count => 20,
burst_limit_window_time => 500}}).
</code>
</section>
<section>
<title>Terminating an Overloaded Handler</title>
<p>It is possible that a handler, even if it can successfully manage peaks
of high load without crashing, can build up a large message queue, or use a
large amount of memory. The overload protection mechanism includes an
automatic termination and restart feature for the purpose of guaranteeing
that a handler does not grow out of bounds. The feature is configured
with the following parameters:</p>
<taglist>
<tag><c>overload_kill_enable</c></tag>
<item>
<p>Value <c>true</c> enables the feature and <c>false</c> disables it.</p>
<p>Defaults to <c>false</c>.</p>
</item>
<tag><c>overload_kill_qlen</c></tag>
<item>
<p>This is the maximum allowed queue length. If the message queue grows
larger than this, the handler process is terminated.</p>
<p>Defaults to <c>20000</c> messages.</p>
</item>
<tag><c>overload_kill_mem_size</c></tag>
<item>
<p>This is the maximum memory size that the handler process is allowed to use.
If the handler grows larger than this, the process is terminated.</p>
<p>Defaults to <c>3000000</c> bytes.</p>
</item>
<tag><c>overload_kill_restart_after</c></tag>
<item>
<p>If the handler is terminated, it restarts automatically after a
delay specified in milliseconds. The value <c>infinity</c> prevents
restarts.</p>
<p>Defaults to <c>5000</c> milliseconds.</p>
</item>
</taglist>
<p>If the handler process is terminated because of overload, it prints
information about it in the log. It also prints information about when a
restart has taken place, and the handler is back in action.</p>
<note>
<p>The sizes of the log events affect the memory needs of the handler.
For information about how to limit the size of log events, see the
<seealso marker="logger_formatter"><c>logger_formatter(3)</c></seealso>
manual page.</p>
</note>
</section>
</section>
<section>
<title>See Also</title>
<p>
<seealso marker="disk_log"><c>disk_log(3)</c></seealso>,
<seealso marker="error_logger"><c>error_logger(3)</c></seealso>,
<seealso marker="logger"><c>logger(3)</c></seealso>,
<seealso marker="logger_disk_log_h"><c>logger_disk_log_h(3)</c></seealso>,
<seealso marker="logger_filters"><c>logger_filters(3)</c></seealso>,
<seealso marker="logger_formatter"><c>logger_formatter(3)</c></seealso>,
<seealso marker="logger_std_h"><c>logger_std_h(3)</c></seealso>,
<seealso marker="sasl:sasl_app"><c>sasl(6)</c></seealso></p>
</section>
</chapter>