Creating Tables

Copy-paste examples of how to create tables with SQL

For the full table creation syntax, see CREATE TABLE.

Column Types

For the full set of column types, see Column Types.

Numeric
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
CREATE TABLE column_types_numeric
(
    b   BOOLEAN,          -- Native "boolean" type
    ti  TINYINT,          -- Native "int8" type
    si  SMALLINT,         -- Native "int16" type
    i   INT,              -- Native "integer" type
    bi  BIGINT,           -- Native "long" type
    ub  UNSIGNED BIGINT,  -- Native "ulong" type
    r   REAL,             -- Native "float" type
    d   DOUBLE,           -- Native "double" type
    dc  DECIMAL(18, 4)    -- Native "decimal" type
)
Text (Simple)
1
2
3
4
5
6
7
CREATE TABLE column_types_text_simple
(
    c_min  VARCHAR(1),    -- Native "char1" type
    c_max  VARCHAR(256),  -- Native "char256" type
    s      VARCHAR,       -- Native "string" type
    sn     VARCHAR(512)   -- Native "string" type
)
Text (Complex)
1
2
3
4
5
6
CREATE TABLE column_types_text_complex
(
    ip  IPV4,  -- Native "ipv4" column property on "string" type
    ui  UUID,  -- Native "uuid" type
    j   JSON   -- Native "json" type
)
Date/Time
1
2
3
4
5
6
7
CREATE TABLE column_types_date_time
(
    td  DATE,      -- Native "date" type
    dt  DATETIME,  -- Native "datetime" type
    tt  TIME,      -- Native "time" type
    ts  TIMESTAMP  -- Native "timestamp" type
)
Binary
1
2
3
4
CREATE TABLE column_types_binary
(
    b  BLOB  -- Native "bytes" type
)
Geospatial
1
2
3
4
5
CREATE TABLE column_types_geospatial
(
    geo  GEOMETRY,  -- Native "wkt" subtype of "string" type
    bw   BLOB(WKT)  -- Native "wkt" subtype of "bytes" type
)
Array
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
CREATE TABLE column_types_array
(
    ab    BOOLEAN[3],  -- Native "array(boolean, 3)" subtype of "string" type
    ai    INT[3],      -- Native "array(int, 3)" subtype of "string" type
    abi   BIGINT[3],   -- Native "array(long, 3)" subtype of "string" type
    avec  VECTOR(3),   -- Native "vector(3)" type
    ar    REAL[3],     -- Native "array(float, 3)" subtype of "string" type
    ad    DOUBLE[3],   -- Native "array(double, 3)" subtype of "string" type
    avar  VARCHAR[3]   -- Native "array(string, 3)" subtype of "string" type
)

Column Features

Any column type can have the nullability modified or have a comment applied. Column properties like dictionary encoding, text searchability, and the initial value modifiers are shown with all applicable column types.

Nullability
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
CREATE TABLE column_features_nullability
(
    b      BOOLEAN       NOT NULL,
    i      INT               NULL,
    r      REAL          NOT NULL,
    d      DOUBLE            NULL,
    v      VARCHAR       NOT NULL,
    c_min  VARCHAR(1)        NULL,
    c_max  VARCHAR(256)  NOT NULL,
    ip     IPV4              NULL,
    ui     UUID          NOT NULL,
    td     DATE              NULL,
    bl     BLOB          NOT NULL,
    geo    GEOMETRY          NULL,
    bw     BLOB(WKT)     NOT NULL,
    ai     INT[3]            NULL,
    avec   VECTOR(3)     NOT NULL
)
Compression
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
CREATE TABLE column_features_compression
(
    b      BOOLEAN(COMPRESS(LZ4(12))),
    i      INT(COMPRESS(NONE)),
    r      REAL(COMPRESS(SNAPPY)),
    d      DOUBLE(COMPRESS(ZSTD(7))),
    dc     DECIMAL(18,4,COMPRESS(LZ4)),
    v      VARCHAR(COMPRESS(ZSTD(3))),
    c_min  VARCHAR(1,COMPRESS(SNAPPY)),
    c_max  VARCHAR(256,COMPRESS(ZSTD)),
    ip     IPV4(COMPRESS(SNAPPY)),
    ui     UUID(COMPRESS(ZSTD)),
    td     DATE(COMPRESS(LZ4(1))),
    bl     BLOB(COMPRESS(NONE)),
    geo    GEOMETRY(COMPRESS(ZSTD(22))),
    bw     BLOB(WKT,COMPRESS(SNAPPY)),
    ai     INT[3](COMPRESS(ZSTD(-1))),
    avec   VECTOR(3,COMPRESS(ZSTD(17)))
)
Dictionary Encoding
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
CREATE TABLE column_features_dictionary_encoding
(
    b      BOOLEAN(DICT),
    ti     TINYINT(DICT),
    si     SMALLINT(DICT),
    i      INT(DICT),
    bi     BIGINT(DICT),
    r      REAL(DICT),
    d      DOUBLE(DICT),
    dc     DECIMAL(10,4,DICT),
    v      VARCHAR(DICT),
    c_min  VARCHAR(1,DICT),
    c_max  VARCHAR(256,DICT),
    ub     UNSIGNED BIGINT(DICT),
    ip     IPV4(DICT),
    ui     UUID(DICT),
    td     DATE(DICT),
    dt     DATETIME(DICT),
    tt     TIME(DICT),
    ts     TIMESTAMP(DICT),
    bl     BLOB(DICT)
)
Text Search
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
CREATE TABLE column_features_text_search
(
    v      VARCHAR(TEXT_SEARCH),
    c_min  VARCHAR(1,TEXT_SEARCH),
    c_max  VARCHAR(256,TEXT_SEARCH),
    td     DATE(TEXT_SEARCH),
    tt     TIME(TEXT_SEARCH),
    dt     DATETIME(TEXT_SEARCH),
    ip     IPV4(TEXT_SEARCH),
    ui     UUID(TEXT_SEARCH)
)
Initial Values
1
2
3
4
5
6
7
8
CREATE TABLE column_features_init_values
(
    ui  UUID(INIT_WITH_UUID),
    td  DATE(INIT_WITH_NOW),
    dt  DATETIME(INIT_WITH_NOW),
    tt  TIME(INIT_WITH_NOW),
    ts  TIMESTAMP(INIT_WITH_NOW)
)
Update Values
1
2
3
4
5
6
7
CREATE TABLE column_features_update_values
(
    td  DATE(UPDATE_WITH_NOW),
    dt  DATETIME(UPDATE_WITH_NOW),
    tt  TIME(UPDATE_WITH_NOW),
    ts  TIMESTAMP(UPDATE_WITH_NOW)
)
Comment
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
CREATE TABLE column_features_comments
(
    b      BOOLEAN      COMMENT 'Native "boolean" type',
    i      INT          COMMENT 'Native "int" type',
    r      REAL         COMMENT 'Native "float" type',
    d      DOUBLE       COMMENT 'Native "double" type',
    v      VARCHAR      COMMENT 'Native "string" type',
    c_min  VARCHAR(1)   COMMENT 'Native "char1" type',
    c_max  VARCHAR(256) COMMENT 'Native "char256" type',
    ip     IPV4         COMMENT 'Native "ipv4" column property on "string" type',
    ui     UUID         COMMENT 'Native "uuid" type',
    td     DATE         COMMENT 'Native "date" type',
    bl     BLOB         COMMENT 'Native "bytes" type',
    geo    GEOMETRY     COMMENT 'Native "wkt" subtype of "string" type',
    bw     BLOB(WKT)    COMMENT 'Native "wkt" subtype of "bytes" type',
    ai     INT[3]       COMMENT 'Native "array(int, 3)" subtype of "string" type',
    avec   VECTOR(3)    COMMENT 'Native "vector(3)" type'
)

Keys

Primary keys, shard keys, & foreign keys, are applicable to most column types. See the respective sections for details.

Primary
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
CREATE TABLE pk
(
    id         TINYINT NOT NULL,
    dept_id    TINYINT,
    manager_id TINYINT,
    first_name VARCHAR(32),
    last_name  VARCHAR(64),
    salary     DECIMAL(18, 4),
    hire_date  DATE,
    PRIMARY KEY (id)
)
Soft Primary
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
CREATE TABLE pk_soft
(
    id         TINYINT NOT NULL,
    dept_id    TINYINT,
    manager_id TINYINT,
    first_name VARCHAR(32),
    last_name  VARCHAR(64),
    salary     DECIMAL(18, 4),
    hire_date  DATE,
    SOFT PRIMARY KEY (id)
)
On-Disk Primary
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
CREATE TABLE pk_disk
(
    id         TINYINT NOT NULL,
    dept_id    TINYINT,
    manager_id TINYINT,
    first_name VARCHAR(32),
    last_name  VARCHAR(64),
    salary     DECIMAL(18, 4),
    hire_date  DATE,
    PRIMARY KEY (id)
)
USING TABLE PROPERTIES (PRIMARY_KEY_TYPE = 'disk')
Shard
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
CREATE TABLE sk
(
    id         TINYINT,
    dept_id    TINYINT,
    manager_id TINYINT,
    first_name VARCHAR(32),
    last_name  VARCHAR(64),
    salary     DECIMAL(18, 4),
    hire_date  DATE,
    SHARD KEY (dept_id)
)
Primary/Shard
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
CREATE TABLE pk_sk
(
    id         TINYINT NOT NULL,
    dept_id    TINYINT NOT NULL,
    manager_id TINYINT,
    first_name VARCHAR(32),
    last_name  VARCHAR(64),
    salary     DECIMAL(18, 4),
    hire_date  DATE,
    PRIMARY KEY (id, dept_id),
    SHARD KEY (dept_id)
)
Foreign
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
CREATE TABLE fk
(
    id         TINYINT,
    dept_id    TINYINT,
    manager_id TINYINT,
    first_name VARCHAR(32),
    last_name  VARCHAR(64),
    salary     DECIMAL(18, 4),
    hire_date  DATE,
    SHARD KEY (dept_id),
    FOREIGN KEY (dept_id) REFERENCES fk_pk(id)
)
Composite Primary
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
CREATE TABLE pk_comp
(
    id         TINYINT NOT NULL,
    dept_id    TINYINT NOT NULL,
    manager_id TINYINT,
    first_name VARCHAR(32),
    last_name  VARCHAR(64),
    salary     DECIMAL(18, 4),
    hire_date  DATE,
    PRIMARY KEY (id, dept_id)
)
Composite Shard
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
CREATE TABLE sk_comp
(
    id         TINYINT,
    dept_id    TINYINT,
    manager_id TINYINT,
    first_name VARCHAR(32),
    last_name  VARCHAR(64),
    salary     DECIMAL(18, 4),
    hire_date  DATE,
    SHARD KEY (id, dept_id)
)
Composite Foreign
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
CREATE TABLE fk_comp
(
    id         TINYINT,
    dept_id    TINYINT,
    manager_id TINYINT,
    first_name VARCHAR(32),
    last_name  VARCHAR(64),
    salary     DECIMAL(18, 4),
    hire_date  DATE,
    SHARD KEY (id, dept_id),
    FOREIGN KEY (id, dept_id) REFERENCES fk_pk_comp(id, dept_id)
)

Indexes

Column & low-cardinality indexes have wide column type applicability; while chunk skip, geospatial, CAGRA, & HNSW indexes are shown with all applicable column types. See the respective sections for details.

Column (Attribute)
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
CREATE TABLE indexes_column
(
    id         TINYINT NOT NULL,
    dept_id    TINYINT,
    manager_id TINYINT,
    first_name VARCHAR(32),
    last_name  VARCHAR(64),
    salary     DECIMAL(18, 4),
    hire_date  DATE
)
INDEX(id)
INDEX(dept_id)
Low-Cardinality
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
CREATE TABLE indexes_low_cardinality
(
    id         TINYINT NOT NULL,
    dept_id    TINYINT NOT NULL,
    manager_id TINYINT,
    first_name VARCHAR(32),
    last_name  VARCHAR(64),
    salary     DECIMAL(18, 4),
    hire_date  DATE
)
LOW CARDINALITY INDEX(dept_id)
LOW CARDINALITY INDEX(manager_id)
Chunk Skip
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
CREATE TABLE indexes_chunk_skip
(
    i      INTEGER,
    bi     BIGINT,
    ub     UNSIGNED BIGINT,
    c_min  VARCHAR(1),
    c_max  VARCHAR(256),
    td     DATE,
    dt     DATETIME,
    tt     TIME,
    ts     TIMESTAMP
)
CHUNK SKIP INDEX(i)
CHUNK SKIP INDEX(bi)
CHUNK SKIP INDEX(ub)
CHUNK SKIP INDEX(c_min)
CHUNK SKIP INDEX(c_max)
CHUNK SKIP INDEX(td)
CHUNK SKIP INDEX(dt)
CHUNK SKIP INDEX(tt)
CHUNK SKIP INDEX(ts)
Geospatial
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
CREATE TABLE indexes_geospatial
(
    low_res_x   REAL,
    low_res_y   REAL,
    high_res_x  DOUBLE,
    high_res_y  DOUBLE,
    geo         GEOMETRY,
    bw          BLOB(WKT)
)
GEOSPATIAL INDEX(low_res_x, low_res_y)
GEOSPATIAL INDEX(high_res_x, high_res_y)
GEOSPATIAL INDEX(geo)
GEOSPATIAL INDEX(bw)
CAGRA
1
2
3
4
5
6
7
8
CREATE TABLE indexes_cagra
(
    v_def  VECTOR(100),
    v_opt  VECTOR(1000)
)
CAGRA INDEX(v_def)
CAGRA INDEX(v_opt)
    WITH OPTIONS (INDEX_OPTIONS = 'itopk_size: 128, graph_build_algo: ivf_pq')
HNSW
1
2
3
4
5
6
7
8
CREATE TABLE indexes_hnsw
(
    v_def  VECTOR(100),
    v_opt  VECTOR(1000)
)
HNSW INDEX(v_def)
HNSW INDEX(v_opt)
    WITH OPTIONS (INDEX_OPTIONS = 'm: 32, ef: 128, ef_construction: 16')

No Error
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
CREATE TABLE IF NOT EXISTS collision_options_no_error
(
    id         TINYINT NOT NULL,
    dept_id    TINYINT,
    manager_id TINYINT,
    first_name VARCHAR(32),
    last_name  VARCHAR(64),
    salary     DECIMAL(18, 4),
    hire_date  DATE
)
Replace
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
CREATE OR REPLACE TABLE collision_options_replace
(
    id         TINYINT NOT NULL,
    dept_id    TINYINT,
    manager_id TINYINT,
    first_name VARCHAR(32),
    last_name  VARCHAR(64),
    salary     DECIMAL(18, 4),
    hire_date  DATE
)

Distribution

See sections on replicated, sharded, & randomly sharded tables for details.

Replicated
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
CREATE REPLICATED TABLE distribution_replicated
(
    id         TINYINT NOT NULL,
    dept_id    TINYINT,
    manager_id TINYINT,
    first_name VARCHAR(32),
    last_name  VARCHAR(64),
    salary     DECIMAL(18, 4),
    hire_date  DATE
)
Sharded (Explicit via Shard Key)
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
CREATE TABLE distribution_sharded_explicit
(
    id         TINYINT NOT NULL,
    dept_id    TINYINT,
    manager_id TINYINT,
    first_name VARCHAR(32),
    last_name  VARCHAR(64),
    salary     DECIMAL(18, 4),
    hire_date  DATE,
    SHARD KEY(id)
)
Sharded (Implicit via Primary Key)
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
CREATE TABLE distribution_sharded_implicit
(
    id         TINYINT NOT NULL,
    dept_id    TINYINT,
    manager_id TINYINT,
    first_name VARCHAR(32),
    last_name  VARCHAR(64),
    salary     DECIMAL(18, 4),
    hire_date  DATE,
    PRIMARY KEY(id)
)
Randomly Sharded
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
CREATE TABLE distribution_random
(
    id         TINYINT,
    dept_id    TINYINT,
    manager_id TINYINT,
    first_name VARCHAR(32),
    last_name  VARCHAR(64),
    salary     DECIMAL(18, 4),
    hire_date  DATE
)

Properties

Memory Only (Temporary)
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
CREATE TEMP TABLE properties_memory_only
(
    id         TINYINT NOT NULL,
    dept_id    TINYINT,
    manager_id TINYINT,
    first_name VARCHAR(32),
    last_name  VARCHAR(64),
    salary     DECIMAL(18, 4),
    hire_date  DATE
)
Chunk Size
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
CREATE TABLE properties_chunk_size
(
    id         TINYINT NOT NULL,
    dept_id    TINYINT,
    manager_id TINYINT,
    first_name VARCHAR(32),
    last_name  VARCHAR(64),
    salary     DECIMAL(18, 4),
    hire_date  DATE
)
USING TABLE PROPERTIES(CHUNK SIZE = 1000000)
Compression
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
CREATE TABLE properties_compression
(
    id         TINYINT NOT NULL,
    dept_id    TINYINT(COMPRESS(LZ4)),
    manager_id TINYINT(COMPRESS(NONE)),
    first_name VARCHAR(32,COMPRESS(SNAPPY)),
    last_name  VARCHAR(64,COMPRESS(LZ4(12))),
    salary     DECIMAL(18,4,COMPRESS(ZSTD(7))),
    hire_date  DATE
)
USING TABLE PROPERTIES(COMPRESSION_CODEC = 'zstd(22)')
TTL
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
CREATE TABLE properties_ttl
(
    id         TINYINT NOT NULL,
    dept_id    TINYINT,
    manager_id TINYINT,
    first_name VARCHAR(32),
    last_name  VARCHAR(64),
    salary     DECIMAL(18, 4),
    hire_date  DATE
)
USING TABLE PROPERTIES(TTL = 15)

Partitions

See section on partitioning for details.

Range
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
CREATE TABLE partition_range_by_year
(
    id           INT NOT NULL,
    customer_id  INT NOT NULL,
    total_price  DECIMAL(10,2),
    purchase_ts  TIMESTAMP NOT NULL
)
PARTITION BY RANGE (YEAR(purchase_ts))
PARTITIONS
(
    order_2014_2016 MIN(2014) MAX(2017),
    order_2017                MAX(2018),
    order_2018                MAX(2019),
    order_2019                MAX(2020)
)
Interval
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
CREATE TABLE partition_interval_by_day
(
    id           INT NOT NULL,
    customer_id  INT NOT NULL,
    total_price  DECIMAL(10,2),
    purchase_ts  TIMESTAMP
)
PARTITION BY INTERVAL (purchase_ts)
PARTITIONS
(
    STARTING AT ('2014-01-01') INTERVAL (INTERVAL '1' DAY)
)
List (Manual)
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
CREATE TABLE partition_list_manual_by_year
(
    id           INT NOT NULL,
    customer_id  INT NOT NULL,
    total_price  DECIMAL(10,2),
    purchase_ts  TIMESTAMP NOT NULL
)
PARTITION BY LIST (YEAR(purchase_ts))
PARTITIONS
(
    order_2014_2016 VALUES (2014, 2015, 2016),
    order_2017      VALUES (2017),
    order_2018      VALUES (2018),
    order_2019      VALUES (2019)
)
List (Manual w/ Multiple Columns)
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
CREATE TABLE partition_list_manual_by_year_and_month
(
    id           INT NOT NULL,
    customer_id  INT NOT NULL,
    total_price  DECIMAL(10,2),
    purchase_ts  TIMESTAMP NOT NULL
)
PARTITION BY LIST (YEAR(purchase_ts), MONTH(purchase_ts))
PARTITIONS
(
    order_2016_0203 VALUES ((2016, 2), (2016, 3)),
    order_2020_03   VALUES ((2020, 3))
)
List (Automatic)
1
2
3
4
5
6
7
8
9
CREATE TABLE partition_list_automatic_by_year_and_month
(
    id           INT NOT NULL,
    customer_id  INT NOT NULL,
    total_price  DECIMAL(10,2),
    purchase_ts  TIMESTAMP NOT NULL
)
PARTITION BY LIST (YEAR(purchase_ts), MONTH(purchase_ts))
AUTOMATIC
Hash
1
2
3
4
5
6
7
8
9
CREATE TABLE partition_hash
(
    id           INT NOT NULL,
    customer_id  INT NOT NULL,
    total_price  DECIMAL(10,2),
    purchase_ts  TIMESTAMP NOT NULL
)
PARTITION BY HASH (YEAR(purchase_ts), MONTH(purchase_ts))
PARTITIONS 10
Series
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
CREATE TABLE partition_series
(
    id           INT NOT NULL,
    customer_id  INT NOT NULL,
    total_price  DECIMAL(10,2),
    purchase_ts  TIMESTAMP NOT NULL,
    SHARD KEY (customer_id)
)
PARTITION BY SERIES (customer_id)
PERCENT_FULL 25

Tier Strategy

See section on the tier strategy clause for details.

Priorities
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
CREATE TABLE tier_strategy_priorities
(
    id           INT NOT NULL,
    name         VARCHAR(128) NOT NULL,
    last_seen    DATETIME
)
TIER STRATEGY
(
    ( ( VRAM 1, RAM 7, PERSIST 5 ) )
)
Disk Cache
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
CREATE TABLE tier_strategy_disk_cache
(
    id           INT NOT NULL,
    name         VARCHAR(128) NOT NULL,
    last_seen    DATETIME
)
TIER STRATEGY
(
    ( ( VRAM 1, RAM 7, DISK0 6, PERSIST 4 ) )
)
Cold Storage
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
CREATE TABLE tier_strategy_cold_storage
(
    id           INT NOT NULL,
    name         VARCHAR(128) NOT NULL,
    last_seen    DATETIME
)
TIER STRATEGY
(
    ( ( VRAM 1, RAM 7, DISK0 6, PERSIST 4, COLD5 5) )
)
Expression-Based
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
CREATE TABLE tier_strategy_expression_based
(
    id           INT NOT NULL,
    name         VARCHAR(128) NOT NULL,
    last_seen    DATETIME
)
TIER STRATEGY
(
    (
        (VRAM 1, RAM 7) WHERE last_seen > TIMESTAMPADD(WEEK, -1, NOW()),
        (RAM 6)
    )
)
Per-Column
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
CREATE TABLE tier_strategy_column_vs_global
(
    id           INT NOT NULL,
    name         VARCHAR(128) NOT NULL,
    last_seen    DATETIME
)
TIER STRATEGY
(
    (
        (VRAM 1, RAM 9) WHERE last_seen > TIMESTAMPADD(DAY, -3, NOW()),
        (VRAM 1, RAM 8, DISK0 7)
    ),
    (
        COLUMNS id, name
            (VRAM 2, RAM 10) WHERE last_seen > TIMESTAMPADD(MINUTE, -1, NOW())
    )
)