System Log

The system_log component stores information about all logged errors and warnings in Home Assistant. All collected information is accessible directly in the frontend, just navigate to the Info section under Developer Tools. In order to not overload Home Assistant with log data, only the 50 last errors and warnings will be stored. Older entries are automatically discarded from the log. It is possible to change the number of stored log entries using the parameter max_entries.

This component is automatically loaded by the frontend (so no need to do anything if you are using the frontend). If you are not doing so, or if you wish to change a parameter, add the following section to your configuration.yaml file:

  max_entries: MAX_ENTRIES

Configuration Variables


(int)(Optional)Number of entries to store (older entries are discarded).

Default value: 50


Service clear

To manually clear the system log, call this service.


Errors and warnings are posted as the event system_log_event, so it is possible to write automations that trigger whenever a warning or error occurs. The following information is included in each event:

Field Description
level Either WARNING or ERROR depending on severity.
source File that triggered the error, e.g., or media_player/
exception Full stack trace if available, otherwise empty string.
message Descriptive message of the error, e.g., “Error handling request”.
timestamp Unix timestamp with as a double, e.g., 1517241010.237416.

Live examples of these events can be found in the Home Assistant log file or by just looking in the system log. An example could, for instance, look like this:

The message (“Unable to find service…”), source ( and level (WARNING) can easily be extracted from the image. Exact timestamp and stack trace is shown if the entry is selected.


Here are some examples using the events posted by system_log.

Counting Number of Warnings

This will create a counter that increases every time a warning is logged:

    name: Warnings
    icon: mdi:alert

  - alias: Count warnings
      platform: event
      event_type: system_log_event
        level: WARNING
      service: counter.increment
      entity_id: counter.warning_counter

Conditional Messages

This automation will create a persistent notification whenever an error or warning is logged that has the word “service” in the message:

  - alias: Create notifications for "service" errors
      platform: event
      event_type: system_log_event
      condition: template
      value_template: '{{ "service" in }}'
      service: persistent_notification.create
        title: Something bad happened
        message: '{{ }}'