Meta functions

These functions provide instance-level information and table, column and partition information including metadata. They are particularly useful for learning widely informative information about your instance and checking the table settings for:

• Designated timestamp column
• Attached, detached, or attachable partitions
• Partition storage size on disk

build

Arguments:

• build() does not require arguments.

Return value:

Returns the current version and hash.

Examples:

build()

build
Build Information: QuestDB 7.3.5, JDK 17.0.7, Commit Hash 460b817b0a3705c5633619a8ef9efb5163f1569c

functions

Arguments:

• functions() does not require arguments.

Return value:

Returns all available database functions.

Examples:

functions()

name	signature	signature_translated	runtime_constant	type
or	or(TT)	or(boolean, boolean)	FALSE	STANDARD
and	and(TT)	and(boolean, boolean)	FALSE	STANDARD
not	not(T)	not(boolean)	FALSE	STANDARD

memory_metrics

Arguments:

• memory_metrics() does not require arguments.

Return value:

Returns granular memory metrics.

Examples:

memory_metrics()

memory_tag	bytes
TOTAL_USED	142624730
RSS	328609792
MMAP_DEFAULT	196728
NATIVE_DEFAULT	256
MMAP_O3	0
NATIVE_O3	96

reader_pool

Arguments:

• reader_pool() does not require arguments.

Return value:

Returns information about the current state of the reader pool in QuestDB. The reader pool is a cache of table readers that are kept open to speed up subsequent reads from the same table. The returned information includes the table name, the ID of the thread that currently owns the reader, the timestamp of the last time the reader was accessed, and the current transaction ID with which the reader is associated.

Examples:

SELECT * FROM reader_pool();

table_name	owner_thread_id	last_access_timestamp	current_txn
sensors	null	2023-12-01T19:28:14.311703Z	1

writer_pool

Arguments:

• writer_pool() does not require arguments.

Return value:

Returns information about the current state of the writer pool in QuestDB. The writer pool is a cache of table writers that are kept open to speed up subsequent writes to the same table. The returned information includes the table name, the ID of the thread that currently owns the writer, the timestamp of the last time the writer was accessed, and the reason for the ownership.

Examples:

SELECT * FROM writer_pool();

table_name	owner_thread_id	last_access_timestamp	ownership_reason
sys.column_versions_purge_log	1	2023-12-01T18:50:03.412468Z	QuestDB system
telemetry_config	1	2023-12-01T18:50:03.470604Z	telemetryConfig
telemetry	1	2023-12-01T18:50:03.464501Z	telemetry
sys.telemetry_wal	1	2023-12-01T18:50:03.467924Z	telemetry
example_table	null	2023-12-01T20:33:33.270984Z	null

current database, schema, or user

current_database(), current_schema(), and current_user() are standard SQL functions that return information about the current database, schema, schemas, and user, respectively.

-- Get the current database

SELECT current_database();

-- Get the current schema

SELECT current_schema();

-- Get the current user

SELECT current_user();

Each of these functions returns a single value, so you can use them in a SELECT statement without any arguments.

tables

tables() or all_tables() returns all tables in the database including table metadata.

Arguments:

• tables() does not require arguments.

Return value:

Returns a table.

Examples:

List all tables

tables();

id	name	designatedTimestamp	partitionBy	maxUncommittedRows	o3MaxLag	walEnabled	directoryName	dedup
1	my_table	ts	DAY	500000	30000000 0	false	my_table	false
2	device_data	null	NONE	10000	30000000	false	device_data	false
3	short_lived	null	HOUR	10000	30000000	false	short_lived (->)	false

All tables in reverse alphabetical order

tables() ORDER BY name DESC;

id	name	designatedTimestamp	partitionBy	maxUncommittedRows	o3MaxLag	walEnabled	directoryName	dedup
2	device_data	null	NONE	10000	30000000	false	device_data	false
1	my_table	ts	DAY	500000	300000000	false	my_table	false
3	short_lived	ts	HOUR	10000	30000000	false	short_lived (->)	false

note

(->) means the table was created using the IN VOLUME clause.

All tables with a daily partitioning strategy

tables() WHERE partitionBy = 'DAY'

id	name	designatedTimestamp	partitionBy	maxUncommittedRows	walEnabled	directoryName	dedup
1	my_table	ts	DAY	500000	true	my_table	false

wal_tables

wal_tables() returns the WAL status for all WAL tables in the database.

Arguments:

• wal_tables() does not require arguments.

Return value:

Returns a table including the following information:

• name - table name
• suspended - suspended status flag
• writerTxn - the last committed transaction in TableWriter
• writerLagTxnCount - the number of transactions that are kept invisible when writing to the table; these transactions will be eventually moved to the table data and become visible for readers
• sequencerTxn - the last committed transaction in the sequencer

Examples:

List all tables

wal_tables();

name	suspended	writerTxn	writerLagTxnCount	sequencerTxn
sensor_wal	false	2	1	4
weather_wal	false	3	0	3
test_wal	true	7	1	9

table_columns

table_columns('tableName') returns the schema of a table.

Arguments:

• tableName is the name of an existing table as a string.

Return value:

Returns a table with the following columns:

• column - name of the available columns in the table
• type - type of the column
• indexed - if indexing is applied to this column
• indexBlockCapacity - how many row IDs to store in a single storage block on disk
• symbolCached - whether this symbol column is cached
• symbolCapacity - how many distinct values this column of symbol type is expected to have
• designated - if this is set as the designated timestamp column for this table
• upsertKey - if this column is a part of UPSERT KEYS list for table deduplication

For more details on the meaning and use of these values, see the CREATE TABLE documentation.

Examples:

Get all columns in a table

table_columns('my_table')

column	type	indexed	indexBlockCapacity	symbolCached	symbolCapacity	designated	upsertKey
symb	SYMBOL	true	1048576	false	256	false	false
price	DOUBLE	false	0	false	0	false	false
ts	TIMESTAMP	false	0	false	0	true	false
s	STRING	false	0	false	0	false	false

Get designated timestamp column

SELECT column, type, designated FROM table_columns('my_table') WHERE designated = true;

column	type	designated
ts	TIMESTAMP	true

Get the count of column types

SELECT type, count() FROM table_columns('my_table');

type	count
SYMBOL	1
DOUBLE	1
TIMESTAMP	1
STRING	1

table_partitions

table_partitions('tableName') returns information for the partitions of a table with the option to filter the partitions.

Arguments:

• tableName is the name of an existing table as a string.

Return value:

Returns a table with the following columns:

• index - INTEGER, index of the partition (NaN when the partition is not attached)
• partitionBy - STRING, one of NONE, HOUR, DAY, WEEK, MONTH and YEAR
• name - STRING, name of the partition, e.g. 2023-03-14, 2023-03-14.detached, 2023-03-14.attachable
• minTimestamp - LONG, min timestamp of the partition (NaN when the table is not partitioned)
• maxTimestamp - LONG, max timestamp of the partition (NaN when the table is not partitioned)
• numRows - LONG, number of rows in the partition
• diskSize - LONG, size of the partition in bytes
• diskSizeHuman - STRING, size of the partition meant for humans to read (same output as function size_pretty)
• readOnly - BOOLEAN, true if the partition is attached via soft link
• active - BOOLEAN, true if the partition is the last partition, and whether we are writing to it (at least one record)
• attached - BOOLEAN, true if the partition is attached
• detached - BOOLEAN, true if the partition is detached (name of the partition will contain the .detached extension)
• attachable - BOOLEAN, true if the partition is detached and can be attached (name of the partition will contain the .attachable extension)

Examples:

Create table my_table

CREATE TABLE my_table AS (

SELECT

rnd_symbol('EURO', 'USD', 'OTHER') symbol,

rnd_double() * 50.0 price,

rnd_double() * 20.0 amount,

to_timestamp('2023-01-01', 'yyyy-MM-dd') + x * 6 * 3600 * 100000L timestamp

FROM long_sequence(700)

), INDEX(symbol capacity 32) TIMESTAMP(timestamp) PARTITION BY WEEK;

Get all partitions from my_table

table_partitions('my_table');

index	partitionBy	name	minTimestamp	maxTimestamp	numRows	diskSize	diskSizeHuman	readOnly	active	attached	detached	attachable
0	WEEK	2022-W52	2023-01-01 00:36:00.0	2023-01-01 23:24:00.0	39	98304	96.0 KiB	false	false	true	false	false
1	WEEK	2023-W01	2023-01-02 00:00:00.0	2023-01-08 23:24:00.0	280	98304	96.0 KiB	false	false	true	false	false
2	WEEK	2023-W02	2023-01-09 00:00:00.0	2023-01-15 23:24:00.0	280	98304	96.0 KiB	false	false	true	false	false
3	WEEK	2023-W03	2023-01-16 00:00:00.0	2023-01-18 12:00:00.0	101	83902464	80.0 MiB	false	true	true	false	false

Get size of a table in disk

SELECT size_pretty(sum(diskSize)) FROM table_partitions('my_table')

size_pretty
80.3 MB

Get active partition of a table

SELECT * FROM table_partitions('my_table') WHERE active = true

index	partitionBy	name	minTimestamp	maxTimestamp	numRows	diskSize	diskSizeHuman	readOnly	active	attached	detached	attachable
3	WEEK	2023-W03	2023-01-16 00:00:00.0	2023-01-18 12:00:00.0	101	83902464	80.0 MiB	false	true	true	false	false

version/pg_catalog.version

version() or pg_catalog.version() returns the supported version of the PostgreSQL Wire Protocol.

Arguments:

• version() or pg_catalog.version() does not require arguments.

Return value:

Returns string.

Examples:

SELECT version();

--The above equals to:

SELECT pg_catalog.version();

version
PostgreSQL 12.3, compiled by Visual C++ build 1914, 64-bit, QuestDB