Aggregate filter plugin

edit
  • Plugin version: v2.9.1
  • Released on: 2019-09-22
  • Changelog

For other versions, see the Versioned plugin docs.

Getting Help

edit

For questions about the plugin, open a topic in the Discuss forums. For bugs or feature requests, open an issue in Github. For the list of Elastic supported plugins, please consult the Elastic Support Matrix.

Description

edit

The aim of this filter is to aggregate information available among several events (typically log lines) belonging to a same task, and finally push aggregated information into final task event.

You should be very careful to set Logstash filter workers to 1 (-w 1 flag) for this filter to work correctly otherwise events may be processed out of sequence and unexpected results will occur.

Example #1

edit
  • with these given logs :
 INFO - 12345 - TASK_START - start
 INFO - 12345 - SQL - sqlQuery1 - 12
 INFO - 12345 - SQL - sqlQuery2 - 34
 INFO - 12345 - TASK_END - end
  • you can aggregate "sql duration" for the whole task with this configuration :
 filter {
   grok {
     match => [ "message", "%{LOGLEVEL:loglevel} - %{NOTSPACE:taskid} - %{NOTSPACE:logger} - %{WORD:label}( - %{INT:duration:int})?" ]
   }

   if [logger] == "TASK_START" {
     aggregate {
       task_id => "%{taskid}"
       code => "map['sql_duration'] = 0"
       map_action => "create"
     }
   }

   if [logger] == "SQL" {
     aggregate {
       task_id => "%{taskid}"
       code => "map['sql_duration'] += event.get('duration')"
       map_action => "update"
     }
   }

   if [logger] == "TASK_END" {
     aggregate {
       task_id => "%{taskid}"
       code => "event.set('sql_duration', map['sql_duration'])"
       map_action => "update"
       end_of_task => true
       timeout => 120
     }
   }
 }
  • the final event then looks like :
{
  "message" => "INFO - 12345 - TASK_END - end message",
  "sql_duration" => 46
}

the field sql_duration is added and contains the sum of all sql queries durations.

Example #2 : no start event

edit
  • If you have the same logs than example #1, but without a start log :
 INFO - 12345 - SQL - sqlQuery1 - 12
 INFO - 12345 - SQL - sqlQuery2 - 34
 INFO - 12345 - TASK_END - end
  • you can also aggregate "sql duration" with a slightly different configuration :
 filter {
   grok {
     match => [ "message", "%{LOGLEVEL:loglevel} - %{NOTSPACE:taskid} - %{NOTSPACE:logger} - %{WORD:label}( - %{INT:duration:int})?" ]
   }

   if [logger] == "SQL" {
     aggregate {
       task_id => "%{taskid}"
       code => "map['sql_duration'] ||= 0 ; map['sql_duration'] += event.get('duration')"
     }
   }

   if [logger] == "TASK_END" {
     aggregate {
       task_id => "%{taskid}"
       code => "event.set('sql_duration', map['sql_duration'])"
       end_of_task => true
       timeout => 120
     }
   }
 }
  • the final event is exactly the same than example #1
  • the key point is the "||=" ruby operator. It allows to initialize sql_duration map entry to 0 only if this map entry is not already initialized

Example #3 : no end event

edit

Third use case: You have no specific end event.

A typical case is aggregating or tracking user behaviour. We can track a user by its ID through the events, however once the user stops interacting, the events stop coming in. There is no specific event indicating the end of the user’s interaction.

In this case, we can enable the option push_map_as_event_on_timeout to enable pushing the aggregation map as a new event when a timeout occurs. In addition, we can enable timeout_code to execute code on the populated timeout event. We can also add timeout_task_id_field so we can correlate the task_id, which in this case would be the user’s ID.

  • Given these logs:
INFO - 12345 - Clicked One
INFO - 12345 - Clicked Two
INFO - 12345 - Clicked Three
  • You can aggregate the amount of clicks the user did like this:
filter {
  grok {
    match => [ "message", "%{LOGLEVEL:loglevel} - %{NOTSPACE:user_id} - %{GREEDYDATA:msg_text}" ]
  }

  aggregate {
    task_id => "%{user_id}"
    code => "map['clicks'] ||= 0; map['clicks'] += 1;"
    push_map_as_event_on_timeout => true
    timeout_task_id_field => "user_id"
    timeout => 600 # 10 minutes timeout
    timeout_tags => ['_aggregatetimeout']
    timeout_code => "event.set('several_clicks', event.get('clicks') > 1)"
  }
}
  • After ten minutes, this will yield an event like:
{
  "user_id": "12345",
  "clicks": 3,
  "several_clicks": true,
    "tags": [
       "_aggregatetimeout"
    ]
}

Example #4 : no end event and tasks come one after the other

edit

Fourth use case : like example #3, you have no specific end event, but also, tasks come one after the other.

That is to say : tasks are not interlaced. All task1 events come, then all task2 events come, …​

In that case, you don’t want to wait task timeout to flush aggregation map.

  • A typical case is aggregating results from jdbc input plugin.
  • Given that you have this SQL query : SELECT country_name, town_name FROM town
  • Using jdbc input plugin, you get these 3 events from :
  { "country_name": "France", "town_name": "Paris" }
  { "country_name": "France", "town_name": "Marseille" }
  { "country_name": "USA", "town_name": "New-York" }
  • And you would like these 2 result events to push them into elasticsearch :
  { "country_name": "France", "towns": [ {"town_name": "Paris"}, {"town_name": "Marseille"} ] }
  { "country_name": "USA", "towns": [ {"town_name": "New-York"} ] }
  • You can do that using push_previous_map_as_event aggregate plugin option :
   filter {
     aggregate {
       task_id => "%{country_name}"
       code => "
         map['country_name'] ||= event.get('country_name')
         map['towns'] ||= []
         map['towns'] << {'town_name' => event.get('town_name')}
         event.cancel()
       "
       push_previous_map_as_event => true
       timeout => 3
     }
   }
  • The key point is that each time aggregate plugin detects a new country_name, it pushes previous aggregate map as a new Logstash event, and then creates a new empty map for the next country
  • When 3s timeout comes, the last aggregate map is pushed as a new event
  • Initial events (which are not aggregated) are dropped because useless (thanks to event.cancel())
  • Last point: if a field is not fulfilled for every event (say "town_postcode" field), the ||= operator will let you to push into aggregate map, the first "not null" value. Example: map['town_postcode'] ||= event.get('town_postcode')

Example #5 : no end event and push events as soon as possible

edit

Fifth use case: like example #3, there is no end event.

Events keep coming for an indefinite time and you want to push the aggregation map as soon as possible after the last user interaction without waiting for the timeout.

This allows to have the aggregated events pushed closer to real time.

A typical case is aggregating or tracking user behaviour.

We can track a user by its ID through the events, however once the user stops interacting, the events stop coming in.

There is no specific event indicating the end of the user’s interaction.

The user interaction will be considered as ended when no events for the specified user (task_id) arrive after the specified inactivity_timeout`.

If the user continues interacting for longer than timeout seconds (since first event), the aggregation map will still be deleted and pushed as a new event when timeout occurs.

The difference with example #3 is that the events will be pushed as soon as the user stops interacting for inactivity_timeout seconds instead of waiting for the end of timeout seconds since first event.

In this case, we can enable the option push_map_as_event_on_timeout to enable pushing the aggregation map as a new event when inactivity timeout occurs.

In addition, we can enable timeout_code to execute code on the populated timeout event.

We can also add timeout_task_id_field so we can correlate the task_id, which in this case would be the user’s ID.

  • Given these logs:
INFO - 12345 - Clicked One
INFO - 12345 - Clicked Two
INFO - 12345 - Clicked Three
  • You can aggregate the amount of clicks the user did like this:
filter {
 grok {
   match => [ "message", "%{LOGLEVEL:loglevel} - %{NOTSPACE:user_id} - %{GREEDYDATA:msg_text}" ]
 }
 aggregate {
   task_id => "%{user_id}"
   code => "map['clicks'] ||= 0; map['clicks'] += 1;"
   push_map_as_event_on_timeout => true
   timeout_task_id_field => "user_id"
   timeout => 3600 # 1 hour timeout, user activity will be considered finished one hour after the first event, even if events keep coming
   inactivity_timeout => 300 # 5 minutes timeout, user activity will be considered finished if no new events arrive 5 minutes after the last event
   timeout_tags => ['_aggregatetimeout']
   timeout_code => "event.set('several_clicks', event.get('clicks') > 1)"
 }
}
  • After five minutes of inactivity or one hour since first event, this will yield an event like:
{
 "user_id": "12345",
 "clicks": 3,
 "several_clicks": true,
   "tags": [
      "_aggregatetimeout"
   ]
}

How it works

edit
  • the filter needs a "task_id" to correlate events (log lines) of a same task
  • at the task beginning, filter creates a map, attached to task_id
  • for each event, you can execute code using event and map (for instance, copy an event field to map)
  • in the final event, you can execute a last code (for instance, add map data to final event)
  • after the final event, the map attached to task is deleted (thanks to end_of_task => true)
  • an aggregate map is tied to one task_id value which is tied to one task_id pattern. So if you have 2 filters with different task_id patterns, even if you have same task_id value, they won’t share the same aggregate map.
  • in one filter configuration, it is recommended to define a timeout option to protect the feature against unterminated tasks. It tells the filter to delete expired maps
  • if no timeout is defined, by default, all maps older than 1800 seconds are automatically deleted
  • all timeout options have to be defined in only one aggregate filter per task_id pattern (per pipeline). Timeout options are : timeout, inactivity_timeout, timeout_code, push_map_as_event_on_timeout, push_previous_map_as_event, timeout_timestamp_field, timeout_task_id_field, timeout_tags
  • if code execution raises an exception, the error is logged and event is tagged _aggregateexception

Use Cases

edit
  • extract some cool metrics from task logs and push them into task final log event (like in example #1 and #2)
  • extract error information in any task log line, and push it in final task event (to get a final event with all error information if any)
  • extract all back-end calls as a list, and push this list in final task event (to get a task profile)
  • extract all http headers logged in several lines to push this list in final task event (complete http request info)
  • for every back-end call, collect call details available on several lines, analyse it and finally tag final back-end call log line (error, timeout, business-warning, …​)
  • Finally, task id can be any correlation id matching your need : it can be a session id, a file path, …​

Aggregate Filter Configuration Options

edit

This plugin supports the following configuration options plus the Common Options described later.

Also see Common Options for a list of options supported by all filter plugins.

 

aggregate_maps_path

edit
  • Value type is string
  • There is no default value for this setting.

The path to file where aggregate maps are stored when Logstash stops and are loaded from when Logstash starts.

If not defined, aggregate maps will not be stored at Logstash stop and will be lost. Must be defined in only one aggregate filter per pipeline (as aggregate maps are shared at pipeline level).

Example:

    filter {
      aggregate {
        aggregate_maps_path => "/path/to/.aggregate_maps"
      }
    }

code

edit
  • This is a required setting.
  • Value type is string
  • There is no default value for this setting.

The code to execute to update aggregated map, using current event.

Or on the contrary, the code to execute to update event, using aggregated map.

Available variables are :

event: current Logstash event

map: aggregated map associated to task_id, containing key/value pairs. Data structure is a ruby Hash

map_meta: meta informations associated to aggregate map. It allows to set a custom timeout or inactivity_timeout. It allows also to get creation_timestamp, lastevent_timestamp and task_id.

When option push_map_as_event_on_timeout=true, if you set map_meta.timeout=0 in code block, then aggregated map is immediately pushed as a new event.

Example:

    filter {
      aggregate {
        code => "map['sql_duration'] += event.get('duration')"
      }
    }

end_of_task

edit
  • Value type is boolean
  • Default value is false

Tell the filter that task is ended, and therefore, to delete aggregate map after code execution.

inactivity_timeout

edit
  • Value type is number
  • There is no default value for this setting.

The amount of seconds (since the last event) after which a task is considered as expired.

When timeout occurs for a task, its aggregate map is evicted.

If push_map_as_event_on_timeout or push_previous_map_as_event is set to true, the task aggregation map is pushed as a new Logstash event.

inactivity_timeout can be defined for each "task_id" pattern.

inactivity_timeout must be lower than timeout.

map_action

edit
  • Value type is string
  • Default value is "create_or_update"

Tell the filter what to do with aggregate map.

"create": create the map, and execute the code only if map wasn’t created before

"update": doesn’t create the map, and execute the code only if map was created before

"create_or_update": create the map if it wasn’t created before, execute the code in all cases

push_map_as_event_on_timeout

edit
  • Value type is boolean
  • Default value is false

When this option is enabled, each time a task timeout is detected, it pushes task aggregation map as a new Logstash event. This enables to detect and process task timeouts in Logstash, but also to manage tasks that have no explicit end event.

push_previous_map_as_event

edit
  • Value type is boolean
  • Default value is false

When this option is enabled, each time aggregate plugin detects a new task id, it pushes previous aggregate map as a new Logstash event, and then creates a new empty map for the next task.

this option works fine only if tasks come one after the other. It means : all task1 events, then all task2 events, etc…​

task_id

edit
  • This is a required setting.
  • Value type is string
  • There is no default value for this setting.

The expression defining task ID to correlate logs.

This value must uniquely identify the task.

Example:

    filter {
      aggregate {
        task_id => "%{type}%{my_task_id}"
      }
    }

timeout

edit
  • Value type is number
  • Default value is 1800

The amount of seconds (since the first event) after which a task is considered as expired.

When timeout occurs for a task, its aggregate map is evicted.

If push_map_as_event_on_timeout or push_previous_map_as_event is set to true, the task aggregation map is pushed as a new Logstash event.

Timeout can be defined for each "task_id" pattern.

timeout_code

edit
  • Value type is string
  • There is no default value for this setting.

The code to execute to complete timeout generated event, when 'push_map_as_event_on_timeout' or 'push_previous_map_as_event' is set to true. The code block will have access to the newly generated timeout event that is pre-populated with the aggregation map.

If 'timeout_task_id_field' is set, the event is also populated with the task_id value

Example:

    filter {
      aggregate {
        timeout_code => "event.set('state', 'timeout')"
      }
    }

timeout_tags

edit
  • Value type is array
  • Default value is []

Defines tags to add when a timeout event is generated and yield

Example:

    filter {
      aggregate {
        timeout_tags => ["aggregate_timeout"]
      }
    }

timeout_task_id_field

edit
  • Value type is string
  • There is no default value for this setting.

This option indicates the timeout generated event’s field where the current "task_id" value will be set. This can help to correlate which tasks have been timed out.

By default, if this option is not set, task id value won’t be set into timeout generated event.

Example:

    filter {
      aggregate {
        timeout_task_id_field => "task_id"
      }
    }

timeout_timestamp_field

edit
  • Value type is string
  • There is no default value for this setting.

By default, timeout is computed using system time, where Logstash is running.

When this option is set, timeout is computed using event timestamp field indicated in this option. It means that when a first event arrives on aggregate filter and induces a map creation, map creation time will be equal to this event timestamp. Then, each time a new event arrives on aggregate filter, event timestamp is compared to map creation time to check if timeout happened.

This option is particularly useful when processing old logs with option push_map_as_event_on_timeout => true. It lets to generate aggregated events based on timeout on old logs, where system time is inappropriate.

Warning : so that this option works fine, it must be set on first aggregate filter.

Example:

    filter {
      aggregate {
        timeout_timestamp_field => "@timestamp"
      }
    }

Common Options

edit

The following configuration options are supported by all filter plugins:

add_field

edit
  • Value type is hash
  • Default value is {}

If this filter is successful, add any arbitrary fields to this event. Field names can be dynamic and include parts of the event using the %{field}.

Example:

    filter {
      aggregate {
        add_field => { "foo_%{somefield}" => "Hello world, from %{host}" }
      }
    }
    # You can also add multiple fields at once:
    filter {
      aggregate {
        add_field => {
          "foo_%{somefield}" => "Hello world, from %{host}"
          "new_field" => "new_static_value"
        }
      }
    }

If the event has field "somefield" == "hello" this filter, on success, would add field foo_hello if it is present, with the value above and the %{host} piece replaced with that value from the event. The second example would also add a hardcoded field.

add_tag

edit
  • Value type is array
  • Default value is []

If this filter is successful, add arbitrary tags to the event. Tags can be dynamic and include parts of the event using the %{field} syntax.

Example:

    filter {
      aggregate {
        add_tag => [ "foo_%{somefield}" ]
      }
    }
    # You can also add multiple tags at once:
    filter {
      aggregate {
        add_tag => [ "foo_%{somefield}", "taggedy_tag"]
      }
    }

If the event has field "somefield" == "hello" this filter, on success, would add a tag foo_hello (and the second example would of course add a taggedy_tag tag).

enable_metric

edit
  • Value type is boolean
  • Default value is true

Disable or enable metric logging for this specific plugin instance by default we record all the metrics we can, but you can disable metrics collection for a specific plugin.

  • Value type is string
  • There is no default value for this setting.

Add a unique ID to the plugin configuration. If no ID is specified, Logstash will generate one. It is strongly recommended to set this ID in your configuration. This is particularly useful when you have two or more plugins of the same type, for example, if you have 2 aggregate filters. Adding a named ID in this case will help in monitoring Logstash when using the monitoring APIs.

    filter {
      aggregate {
        id => "ABC"
      }
    }

periodic_flush

edit
  • Value type is boolean
  • Default value is false

Call the filter flush method at regular interval. Optional.

remove_field

edit
  • Value type is array
  • Default value is []

If this filter is successful, remove arbitrary fields from this event. Fields names can be dynamic and include parts of the event using the %{field} Example:

    filter {
      aggregate {
        remove_field => [ "foo_%{somefield}" ]
      }
    }
    # You can also remove multiple fields at once:
    filter {
      aggregate {
        remove_field => [ "foo_%{somefield}", "my_extraneous_field" ]
      }
    }

If the event has field "somefield" == "hello" this filter, on success, would remove the field with name foo_hello if it is present. The second example would remove an additional, non-dynamic field.

remove_tag

edit
  • Value type is array
  • Default value is []

If this filter is successful, remove arbitrary tags from the event. Tags can be dynamic and include parts of the event using the %{field} syntax.

Example:

    filter {
      aggregate {
        remove_tag => [ "foo_%{somefield}" ]
      }
    }
    # You can also remove multiple tags at once:
    filter {
      aggregate {
        remove_tag => [ "foo_%{somefield}", "sad_unwanted_tag"]
      }
    }

If the event has field "somefield" == "hello" this filter, on success, would remove the tag foo_hello if it is present. The second example would remove a sad, unwanted tag as well.