Fluent Bit and SQL

Stream processing in Fluent Bit uses SQL to perform record queries.

For more information, see the stream processing README file.

Statements

Use the following SQL statements in Fluent Bit.

SELECT

SELECT results_statement
  FROM STREAM:stream_name | TAG:match_rule
  [WINDOW TUMBLING (integer SECOND) | WINDOW HOPPING (integer SECOND, ADVANCE BY integer SECOND)]
  [WHERE condition]
  [GROUP BY groupby]

Groups keys from records that originate from a specified stream, or from records that match a specific tag pattern.

A SELECT statement not associated with stream creation will send the results to the standard output interface, which can be helpful for debugging purposes.

You can filter the results of this query by applying a condition by using a WHERE statement. For information about the WINDOW and GROUP BY statements, see Aggregation functions.

Examples [#select-examples]

Selects all keys from records that originate from a stream called apache:

SELECT * FROM STREAM:apache;

Selects the code key from records with tags whose name begins with apache:

SELECT code AS http_status FROM TAG:'apache.*';

CREATE STREAM

CREATE STREAM stream_name
  [WITH (property_name=value, [...])]
  AS select_statement

Creates a new stream of data using the results from a SELECT statement. If the Tag property in the WITH statement is set, this new stream can optionally be re-ingested into the Fluent Bit pipeline.

Examples [#create-stream-examples]

Creates a new stream called hello_ from a stream called apache:

CREATE STREAM hello AS SELECT * FROM STREAM:apache;

Creates a new stream called hello for all records whose original tag name begins with apache:

CREATE STREAM hello AS SELECT * FROM TAG:'apache.*';

Aggregation functions

You can use aggregation functions in the results_statement on keys, which lets you perform data calculation on groups of records. These groups are determined by the WINDOW key. If WINDOW is unspecified, aggregation functions are applied to the current buffer of records received, which might have a non-deterministic number of elements. You can also apply aggregation functions to records in a window of a specific time interval.

Fluent Bit supports two window types:

• Tumbling window (WINDOW TUMBLING): Non-overlapping windows. A window size of 5 performs aggregation on records during a five-second interval, then starts a fresh window for the next interval.

• Hopping window (WINDOW HOPPING): A sliding window with a configurable advance step. For example, WINDOW HOPPING (10 SECOND, ADVANCE BY 2 SECOND) maintains a 10-second window that advances every 2 seconds, so consecutive windows share overlapping records.

Additionally, you can use the GROUP BY statement to group results by one or more keys with matching values.

AVG

SELECT AVG(size) FROM STREAM:apache WHERE method = 'POST' ;

Calculates the average size of POST requests.

COUNT

SELECT host, COUNT(*) FROM STREAM:apache WINDOW TUMBLING (5 SECOND) GROUP BY host;

Counts the number of records in a five-second tumbling window, grouped by host IP addresses.

MIN

SELECT MIN(key) FROM STREAM:apache;

Returns the minimum value of a key in a set of records.

MAX

SELECT MAX(key) FROM STREAM:apache;

Returns the maximum value of a key in a set of records.

SUM

SELECT SUM(key) FROM STREAM:apache;

Calculates the sum of all values of a key in a set of records.

TIMESERIES_FORECAST

SELECT TIMESERIES_FORECAST(num, 30) FROM STREAM:apache WINDOW TUMBLING (5 SECOND);

Uses linear regression to predict the future value of a key. The first argument is the key to forecast and the second argument is the number of seconds into the future to project. Requires a WINDOW to accumulate the data points used for the regression.

WINDOW HOPPING example

SELECT host, COUNT(*) FROM STREAM:apache WINDOW HOPPING (10 SECOND, ADVANCE BY 2 SECOND) GROUP BY host;

Counts records per host using a 10-second hopping window that advances every 2 seconds. Each output overlaps with the previous window, unlike a tumbling window.

Time functions

Use time functions to add a new key with time data into a record.

NOW

SELECT NOW() FROM STREAM:apache;

Adds the current system time to a record using the format %Y-%m-%d %H:%M:%S. Output example: 2019-03-09 21:36:05.

UNIX_TIMESTAMP

SELECT UNIX_TIMESTAMP() FROM STREAM:apache;

Adds the current Unix time to a record. Output example: 1552196165.

Record functions

Use record functions to append new keys to a record using values from the record's context.

RECORD_TAG

SELECT RECORD_TAG() FROM STREAM:apache;

Append tag string associated to the record as a new key.

RECORD_TIME

SELECT RECORD_TIME() FROM STREAM:apache;

Appends the record's timestamp as a new key in double format (seconds.nanoseconds). Output example: 1552196165.705683.

WHERE condition

Similar to conventional SQL statements, Fluent Bit supports the WHERE condition. You can use this condition in both keys and subkeys. For example:

SELECT AVG(size) FROM STREAM:apache WHERE method = 'POST' AND status = 200;

You can confirm whether a key exists in a record by using the record-specific function @record.contains:

SELECT MAX(key) FROM STREAM:apache WHERE @record.contains(key);

To determine if the value of a key is NULL:

SELECT MAX(key) FROM STREAM:apache WHERE key IS NULL;

Or similar:

SELECT * FROM STREAM:apache WHERE user IS NOT NULL;