query-comment
query-comment: string
The query-comment
configuration also accepts a dictionary input, like so:
models:
my_dbt_project:
+materialized: table
query-comment:
comment: string
append: true | false
job-label: true | false # BigQuery only
Definition
A string to inject as a comment in each query that dbt runs against your database. This comment can attribute SQL statements to specific dbt resources like models and tests.
The query-comment
configuration can also call a macro that returns a string.
Default
By default, dbt will insert a JSON comment at the top of your query containing the information including the dbt version, profile and target names, and node ids for the resources it runs. For example:
/* {"app": "dbt", "dbt_version": "0.15.0rc2", "profile_name": "debug",
"target_name": "dev", "node_id": "model.dbt2.my_model"} */
create view analytics.analytics.orders as (
select ...
);
Using the dictionary syntax
The dictionary syntax includes two keys:
comment
(optional, for more information, refer to the default section): The string to be injected into a query as a comment.append
(optional, default=false
): Whether a comment should be appended (added to the bottom of a query) or not (i.e. added to the top of a query). By default, comments are added to the top of queries (i.e.append: false
).
This syntax is useful on databases like Snowflake which remove leading SQL comments.
Examples
Prepend a static comment
The following example injects a comment that reads /* executed by dbt */
into the header of the SQL queries that dbt runs.
query-comment: "executed by dbt"
Example output:
/* executed by dbt */
select ...
Disable query comments
query-comment:
Or:
query-comment: null
Prepend a dynamic comment
The following example injects a comment that varies based on the configured user
specified in the active dbt target.
query-comment: "run by {{ target.user }} in dbt"
Example output:
/* run by drew in dbt */
select ...
Append the default comment
The following example uses the dictionary syntax to append (rather than prepend) the default comment.
Note that the comment:
field is omitted to allow the default to be appended.
query-comment:
append: True
Example output:
select ...
/* {"app": "dbt", "dbt_version": "0.16.`0rc2`", "profile_name": "debug", "target_name": "dev", "node_id": "model.dbt2.my_model"} */
;
BigQuery: include query comment items as job labels
If query-comment.job-label
is set to true, dbt will include the query comment items, if a dictionary, or the comment string, as job labels on the query it executes. These will be included in addition to labels specified in the BigQuery-specific config.
query-comment:
job-label: True
Append a custom comment
The following example uses the dictionary syntax to append (rather than prepend) a comment that varies based on the configured user
specified in the active dbt target.
query-comment:
comment: "run by {{ target.user }} in dbt"
append: True
Example output:
select ...
/* run by drew in dbt */
;
Intermediate: Use a macro to generate a comment
The query-comment
config can reference macros in your dbt project. Simply create a macro with any name (query_comment
is a good start!) in your macros
directory, like so:
{% macro query_comment() %}
dbt {{ dbt_version }}: running {{ node.unique_id }} for target {{ target.name }}
{% endmacro %}
Then call the macro in your dbt_project.yml
file. Make sure you quote the macro to avoid the YAML parser from trying to interpret the {
as the start of a dictionary.
query-comment: "{{ query_comment() }}"
Advanced: Use a macro to generate a comment
The following example shows a JSON query comment which can be parsed to understand the performance characteristics of your dbt project.
{% macro query_comment(node) %}
{%- set comment_dict = {} -%}
{%- do comment_dict.update(
app='dbt',
dbt_version=dbt_version,
profile_name=target.get('profile_name'),
target_name=target.get('target_name'),
) -%}
{%- if node is not none -%}
{%- do comment_dict.update(
file=node.original_file_path,
node_id=node.unique_id,
node_name=node.name,
resource_type=node.resource_type,
package_name=node.package_name,
relation={
"database": node.database,
"schema": node.schema,
"identifier": node.identifier
}
) -%}
{% else %}
{%- do comment_dict.update(node_id='internal') -%}
{%- endif -%}
{% do return(tojson(comment_dict)) %}
{% endmacro %}
As above, call this macro as follows:
query-comment: "{{ query_comment(node) }}"
Compilation context
The following context variables are available when generating a query comment:
Context Variable | Description |
---|---|
dbt_version | The version of dbt being used. For details about release versioning, refer to Versioning. |
env_var | See env_var |
modules | See modules |
run_started_at | When the dbt invocation began |
invocation_id | A unique ID for the dbt invocation |
fromjson | See fromjson |
tojson | See tojson |
log | See log |
var | See var |
target | See target |
connection_name | A string representing the internal name for the connection. This string is generated by dbt. |
node | A dictionary representation of the parsed node object. Use node.unique_id , node.database , node.schema , and so on. |
Note: The var()
function in query-comment
macros only access variables passed through the --vars
argument in the CLI. Variables defined in the vars block of your dbt_project.yml
are not accessible when generating query comments.