Viewing trace data

You can view trace data either through Snowsight or by querying the event table in which trace data is stored.

Note

Before you can begin using trace data, you must enable telemetry data collection.

View trace entries in Snowsight

Note

Snowsight support for logging and tracing is a preview feature available to all accounts.

You can use Snowsight to view trace data captured in the event table.

  1. Sign in to Snowsight.

  2. In the navigation menu, select Monitoring » Traces and Logs.

  3. In the Traces & Logs page, you can view trace entries with the following columns:

    Column

    Description

    Date

    Date on which the entry was recorded.

    Duration

    Length of time from the trace’s start to its end.

    Trace Name

    The name of the executable generating the event. Retrieved from the RESOURCE_ATTRIBUTES column snow.executable.name value.

    Status

    Error if the trace reported errors; otherwise, Success.

    Spans

    Number of spans in the trace.

  4. In the Traces & Logs page, you can perform the following actions:

    • To filter the displayed rows, use the drop-down menus at the top of the page. You can filter by the following characteristics:

      • Date range during which the entry was recorded

      • Status of the trace, such as Success or Error

      • Database on which the trace was run

    • To sort rows, select the name of the column by which you want to sort.

  5. To view more detailed information about an entry in its Trace Details page, select the entry’s row.

  6. In the Traces Details page, you can view a list of spans.

    A span object contains trace events. For more information, see How Snowflake represents trace events.

    • To filter the displayed rows use the drop-down menu at the top of the page. You can filter by Span Type: UDF, procedure, or Streamlit.

    • To view a legend describing data shown in the rows, select the Legend dropdown, and then select the legend you want to see.

    • To view more detailed information about an entry, select the entry’s row.

      On this panel, you can view more information stored in the event table. The following tables describe values in the panel:

      Details tab

      Detail

      Description

      Trace ID

      A unique identifier for calls made from a query. Retrieved from the TRACE column trace_id value. For more information, see Trace value.

      Span ID

      A unique identifier tied to the threading model. Retrieved from the TRACE column span_id value. For more information, see Trace value.

      Scope

      Namespace of code emitting the event. Retrieved from the SCOPE column.

      Duration

      The span’s duration from start to finish. For more information, see For SPAN RECORD_TYPE.

      Name

      The span’s name. Retrieved from the RECORD column name value.

      Parent Span ID

      Unique ID of the span containing the selected span.

      Status Code

      The span’s status code. Retrieved from the RECORD column status value.

      Other attributes

      Attributes and values added by user code.

      Query ID

      ID of the query that initiated the trace. Retrieved from the RESOURCE_ATTRIBUTES column snow.query.id value.

      Name

      The name of the executable generating the event. Retrieved from the RESOURCE_ATTRIBUTES column snow.executable.name value.

      Type

      The type of executable that generated the event. Retrieved from the RESOURCE_ATTRIBUTES column snow.executable.type value.

      User

      The name of the user executing the function or procedure. For a Streamlit app, the name of the user who was viewing the app for a given event. Retrieved from the RESOURCE_ATTRIBUTES column db.user value.

      Owner

      The name of the role with OWNERSHIP privilege for the executable. Retrieved from the RESOURCE_ATTRIBUTES column snow.owner.name value.

      Role

      The name of the primary role in the session. Retrieved from the RESOURCE_ATTRIBUTES column snow.session.role.primary.name value.

      Warehouse

      The name of the warehouse running the query generating the event. Retrieved from the RESOURCE_ATTRIBUTES column snow.warehouse.name value.

      Database

      The name of the database containing the executable. Retrieved from the RESOURCE_ATTRIBUTES column snow.database.name value.

      Schema

      The name of the schema containing the executable. Retrieved from the RESOURCE_ATTRIBUTES column snow.schema.name value.

      Span Events tab

      Displays data recorded for trace events. For more information, see Event data recorded.

      Related Metrics tab

      Displays charts illustrating CPU and memory metrics for resource consumption by Snowpark Python stored procedures and UDFs. Metrics associated with the UDF are for a specific query. If you select a UDF span from the list, the metrics are related to one or more UDF spans for the same query. If you select a procedure from the list, you will see procedure metrics for a single span.

      Logs tab

      Displays the value logged by the event. Retrieved from the VALUE column.

Query the event table for trace entries

An event table has a set of predefined columns that capture information about the logged messages, including:

  • The timestamp when a span began.

  • The timestamp when the event was created.

  • The type of data recorded, such as whether the data is for a span or span event.

  • The name of the span or event.

  • Attributes, if any, associated with the span or event.

For reference information about event table columns, see Event table columns.

Trace data query example

The following sections illustrate with example data how you can query the event table for trace data.

Collected data

Output in the following example shows content from a selected subset of columns from an event table after trace data has been captured for three separate handlers written in Python.

For reference information on event table columns that collect trace data, see Data for trace events.

-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
| TIMESTAMP          | START_TIMESTAMP    | RESOURCE_ATTRIBUTES   | RECORD_TYPE | RECORD                                                                                                  | RECORD_ATTRIBUTES                                                           |
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
| 2023-04-20 0:45:49 | 2023-04-20 0:45:49 | **See excerpt below** | SPAN        | { "kind": "SPAN_KIND_INTERNAL", "name": "digits_of_number", "status": { "code": "STATUS_CODE_UNSET" } } |                                                                             |
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
| 2023-04-20 0:45:49 |                    |                       | SPAN_EVENT  | { "name": "test_udtf_init" }                                                                            |                                                                             |
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
| 2023-04-20 0:45:49 |                    |                       | SPAN_EVENT  | { "name": "test_udtf_process" }                                                                         | { "input": "42" }                                                           |
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
| 2023-04-20 0:45:49 |                    |                       | SPAN_EVENT  | { "name": "test_udtf_end_partition" }                                                                   |                                                                             |
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
| 2023-04-20 0:46:00 | 2023-04-20 0:46:00 |                       | SPAN        | { "kind": "SPAN_KIND_INTERNAL", "name": "times_two", "status": { "code": "STATUS_CODE_UNSET" } }        | { "example.func.times_two": "begin", "example.func.times_two.response": 8 } |
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
| 2023-04-20 0:46:00 |                    |                       | SPAN_EVENT  | { "name": "event_without_attributes" }                                                                  |                                                                             |
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
| 2023-04-20 0:46:00 |                    |                       | SPAN_EVENT  | { "name": "event_with_attributes" }                                                                     | { "example.key1": "value1", "example.key2": "value2" }                      |
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
| 2023-04-20 0:46:08 | 2023-04-20 0:46:08 |                       | SPAN        | { "kind": "SPAN_KIND_INTERNAL", "name": "do_tracing", "status": { "code": "STATUS_CODE_UNSET" } }       | { "example.proc.do_tracing": "begin" }                                      |
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
| 2023-04-20 0:46:08 |                    |                       | SPAN_EVENT  | { "name": "event_with_attributes" }                                                                     | { "example.key1": "value1", "example.key2": "value2" }                      |
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

RESOURCE_ATTRIBUTES Excerpts

The following JSON excerpts contain two of the attributes included in the RESOURCE_ATTRIBUTES column for each of the three handlers whose data is included in the preceding output. The SELECT statement following these excerpts selects values from these attributes.

The RESOURCE_ATTRIBUTES column contains data about the event’s source. For reference information, see RESOURCE_ATTRIBUTES column.

{
  ...
  "snow.executable.name": "DIGITS_OF_NUMBER(INPUT NUMBER):TABLE: (RESULT NUMBER)",
  "snow.executable.type": "FUNCTION",
  ...
}

{
  ...
  "snow.executable.name": "TIMES_TWO(X NUMBER):NUMBER(38,0)",
  "snow.executable.type": "FUNCTION",
  ...
}

{
  ...
  "snow.executable.name": "DO_TRACING():VARIANT",
  "snow.executable.type": "PROCEDURE",
  ...
}
Copy

Query with SELECT statement

When querying for data, you can select attribute values within a column by using bracket notation, as in the following form:

COLUMN_NAME['attribute_name']
Copy

Code in the example below queries the preceding table with the intention of isolating data related to the DIGITS_OF_NUMBER function.

SET EVENT_TABLE_NAME='my_db.public.my_events';

SELECT
  TIMESTAMP as time,
  RESOURCE_ATTRIBUTES['snow.executable.name'] as handler_name,
  RESOURCE_ATTRIBUTES['snow.executable.type'] as handler_type,
  RECORD['name'] as event_name,
  RECORD_ATTRIBUTES as attributes
FROM
  IDENTIFIER($event_table_name)
WHERE
  RECORD_TYPE = 'SPAN_EVENT'
  AND HANDLER_NAME LIKE 'DIGITS_OF_NUMBER%';
Copy

Query results

Output in the following example illustrates the query’s result.

-------------------------------------------------------------------------------------------------------------------------------------------
| TIME               | HANDLER_NAME                                          | HANDLER_TYPE | EVENT_NAME              | ATTRIBUTES        |
-------------------------------------------------------------------------------------------------------------------------------------------
| 2023-04-20 0:45:49 | DIGITS_OF_NUMBER(INPUT NUMBER):TABLE: (RESULT NUMBER) | FUNCTION     | test_udtf_init          |                   |
-------------------------------------------------------------------------------------------------------------------------------------------
| 2023-04-20 0:45:49 | DIGITS_OF_NUMBER(INPUT NUMBER):TABLE: (RESULT NUMBER) | FUNCTION     | test_udtf_process       | { "input": "42" } |
-------------------------------------------------------------------------------------------------------------------------------------------
| 2023-04-20 0:45:49 | DIGITS_OF_NUMBER(INPUT NUMBER):TABLE: (RESULT NUMBER) | FUNCTION     | test_udtf_end_partition |                   |
-------------------------------------------------------------------------------------------------------------------------------------------
Language: English