Create a Table

On this page Carat arrow pointing down

This page provides best-practice guidance on creating tables, with some simple examples based on Cockroach Labs's fictional vehicle-sharing company, MovR.

Tip:

For detailed reference documentation on the CREATE TABLE statement, including additional examples, see the CREATE TABLE syntax page.

Before you begin

Before reading this page, do the following:

Create a table

Tables are the logical objects in a cluster that store data sent from your application's persistence layer. Tables organize records of data in rows and columns.

To create a table, use a CREATE TABLE statement, following the best practices that we have listed in the following sections:

After reviewing the best practices in each section, see the example provided in that section.

Name a table

Naming a table is the first step in table creation.

CREATE TABLE statements generally take the form:

CREATE TABLE {schema_name}.{table_name} (
  {elements}
  );
Parameter Description
{schema_name} The name of the user-defined schema.
{table_name} The name of the table.
{elements} A comma-separated list of table elements, such as column definitions.

For an example, see below.

Table naming best practices

Here are some best practices to follow when naming tables:

  • Use a fully-qualified name (i.e., CREATE TABLE database_name.schema_name.table_name). If you do not specify the database name, CockroachDB will use the SQL session's current database (defaultdb, by default). If you do not specify the user-defined schema in the table name, CockroachDB will create the table in the preloaded public schema.

  • Use a table name that reflects the contents of the table. For example, for a table containing information about orders, you could use the name orders (as opposed to naming the table something like table1).

Table naming example

Suppose you want to create a table to store information about users of the MovR platform, and you want the SQL user max to manage that table.

Create an empty .sql file for max:

icon/buttons/copy
$ touch max_init.sql

This file will initialize the objects in the max_schema user-defined schema that you created in Create a Schema, starting with a users table.

In a text editor, open max_init.sql, and add an empty CREATE TABLE statement for the users table:

icon/buttons/copy
CREATE TABLE movr.max_schema.users (
);

Next, define the table's columns.

Define columns

Column definitions give structure to a table by separating the values in each row into columns of a single data type.

Column definitions generally take the following form:

{column_name} {DATA_TYPE} {column_qualification}
Parameter Description
{column_name} The name of the column.
{DATA_TYPE} The data type of the row values in the column.
{column_qualification} Some column qualification, such as a column-level constraint, or a computed column clause.

For examples, see below.

Column definition best practices

Here are some best practices to follow when defining table columns:

  • Review the supported column data types, and select the appropriate type for the data you plan to store in a column, following the best practices listed on the data type's reference page.

  • Use column data types with a fixed size limit, or set a maximum size limit on column data types of variable size (e.g., VARBIT(n)). Values exceeding 1MB can lead to write amplification and cause significant performance degradation.

  • Review the primary key best practices and examples, decide if you need to define any dedicated primary key columns.

  • Review the best practices and examples for adding additional constraints, and decide if you need to add any additional constraints to your columns.

Warning:

We strongly recommend adding size limits to all indexed columns, which includes columns in primary keys.

Values exceeding 1 MiB can lead to storage layer write amplification and cause significant performance degradation or even crashes due to OOMs (out of memory errors).

To add a size limit using CREATE TABLE:

icon/buttons/copy
CREATE TABLE name (first STRING(100), last STRING(100));

To add a size limit using ALTER TABLE ... ALTER COLUMN:

icon/buttons/copy
SET enable_experimental_alter_column_type_general = true;
icon/buttons/copy
ALTER TABLE name ALTER first TYPE STRING(99);

Column definition examples

In the max_init.sql file, add a few column definitions to the users table's CREATE TABLE statement, for user names and email addresses:

icon/buttons/copy
CREATE TABLE movr.max_schema.users (
    first_name STRING,
    last_name STRING,
    email STRING
);

All of the columns shown above use the STRING data type, meaning that any value in any of the columns must be of the data type STRING.

CockroachDB supports a number of other column data types, including DECIMAL, INT, TIMESTAMP, UUID, and enumerated data types and spatial data types. We recommend that you review the supported types, and create columns with data types that correspond to the types of data that you intend to persist to the cluster from your application.

Let's add another example table to our max_schema schema, with more column data types.

As a vehicle-sharing platform, MovR needs to store data about its vehicles. In max_init.sql, add a CREATE TABLE statement for a vehicles table, under the CREATE TABLE statement for users. This table should probably include information about the type of vehicle, when it was created, what its availability is, and where it is located:

icon/buttons/copy
CREATE TABLE movr.max_schema.vehicles (
      id UUID,
      type STRING,
      creation_time TIMESTAMPTZ,
      available BOOL,
      last_location STRING
  );

This table includes a few more data types than the users table:

The rest of the columns are STRING-typed.

Note that values in the type column will likely only be STRING values from a fixed list of values. Specifically, the vehicle type can only be one of the vehicle types supported by the MovR platform (e.g., a bike, a scooter, or a skateboard). For values like this, we recommend using a user-defined, enumerated type.

To create a user-defined type, use a CREATE TYPE statement. For example, above the CREATE TABLE statement for the vehicles table, add the following statements:

icon/buttons/copy
CREATE TYPE movr.max_schema.vtype AS ENUM ('bike', 'scooter', 'skateboard');
Tip:

For detailed reference documentation on the CREATE TYPE statement, including additional examples, see the CREATE TYPE syntax page.
For detailed reference documentation on enumerated data types, including additional examples, see ENUM.

You can then use movr.max_schema.vtype as the type column's data type:

icon/buttons/copy
CREATE TABLE movr.max_schema.vehicles (
      id UUID,
      type movr.max_schema.vtype,
      creation_time TIMESTAMPTZ,
      available BOOL,
      last_location STRING
  );

Only values in the set of movr.max_schema.vtype values will be allowed in the type column.

The users and vehicles tables now have syntactically valid column definitions. As a best practice, you should explicitly select primary key columns and add any additional constraints before executing the CREATE TABLE statements.

Select primary key columns

A primary key is a column, or set of columns, whose values uniquely identify rows of data. Every table requires a primary key.

Primary keys are defined in CREATE TABLE statements with the PRIMARY KEY column constraint. The PRIMARY KEY constraint requires that all the constrained column(s) contain only unique and non-NULL values.

When a table is created, CockroachDB creates an index (called the primary index and named {tbl}_pkey) on the column(s) constrained by the PRIMARY KEY constraint. CockroachDB uses this index to find rows in a table more efficiently.

To add a single column to a primary key, add the PRIMARY KEY keyword to the end of the column definition. To add multiple columns to a primary key (i.e., to create a composite primary key), add a separate CONSTRAINT "primary" PRIMARY KEY clause after the column definitions in the CREATE TABLE statement.

For examples, see below.

Tip:

For detailed reference documentation on the PRIMARY KEY constraint, including additional examples, see the PRIMARY KEY constraint page.

Primary key best practices

Here are some best practices to follow when selecting primary key columns:

  • Avoid defining primary keys over a single column of sequential data.

    Querying a table with a primary key on a single sequential column (e.g., an auto-incrementing INT column, or a TIMESTAMP value) can result in single-range hot spots that negatively affect performance, or cause transaction contention.

    If you are working with a table that must be indexed on sequential keys, use hash-sharded indexes. For details about the mechanics and performance improvements of hash-sharded indexes in CockroachDB, see our Hash Sharded Indexes Unlock Linear Scaling for Sequential Workloads blog post.

  • Define a primary key for every table.

    If you create a table without defining a primary key, CockroachDB will automatically create a primary key over a hidden, INT-typed column named rowid. By default, sequential, unique identifiers are generated for each row in the rowid column with the unique_rowid() function. The sequential nature of the rowid values can lead to a poor distribution of the data across a cluster, which can negatively affect performance. Furthermore, because you cannot meaningfully use the rowid column to filter table data, the primary key index on rowid does not offer any performance optimization. This means you will always have improved performance by defining a primary key for a table.

    To require an explicitly defined primary key for all tables created in your cluster, set the sql.defaults.require_explicit_primary_keys.enabled cluster setting to true.

  • When possible, define primary key constraints over multiple columns (i.e., use composite primary keys).

    When defining composite primary keys, make sure the data in the first column of the primary key prefix is well-distributed across the nodes in the cluster. To improve the performance of ordered queries, you can add monotonically increasing primary key columns after the first column of the primary key prefix. For an example, see below.

  • For single-column primary keys, use UUID-typed columns with default values randomly-generated, using the gen_random_uuid() SQL function.

    Randomly generating UUID values ensures that the primary key values will be unique and well-distributed across a cluster. For an example, see below.

Warning:

We strongly recommend adding size limits to all indexed columns, which includes columns in primary keys.

Values exceeding 1 MiB can lead to storage layer write amplification and cause significant performance degradation or even crashes due to OOMs (out of memory errors).

To add a size limit using CREATE TABLE:

icon/buttons/copy
CREATE TABLE name (first STRING(100), last STRING(100));

To add a size limit using ALTER TABLE ... ALTER COLUMN:

icon/buttons/copy
SET enable_experimental_alter_column_type_general = true;
icon/buttons/copy
ALTER TABLE name ALTER first TYPE STRING(99);

Note:

Use ALTER ROLE ALL SET {sessionvar} = {val} instead of the sql.defaults.* cluster settings. This allows you to set a default value for all users for any session variable that applies during login, making the sql.defaults.* cluster settings redundant.

Primary key examples

To follow a primary key best practice, the CREATE TABLE statements in max_init.sql for the users and vehicles tables need to explicitly define a primary key.

In the max_init.sql file, add a composite primary key on the first_name and last_name columns of the users table:

icon/buttons/copy
CREATE TABLE movr.max_schema.users (
    first_name STRING,
    last_name STRING,
    email STRING,
    CONSTRAINT "primary" PRIMARY KEY (first_name, last_name)
);

This primary key will uniquely identify rows of user data.

Because PRIMARY KEY constraints imply UNIQUE indexes, only one unique combination of first_name and last_name will be allowed in rows of the users table. Additionally, because PRIMARY KEY implies the NOT NULL constraint, all rows of data inserted into the users table must include values for first_name and last_name.

Primary key columns can also be single columns, if those columns are guaranteed to uniquely identify the row. Their values should also be well-distributed across the cluster.

In the vehicles table definition, add a PRIMARY KEY constraint on the id column:

icon/buttons/copy
CREATE TABLE movr.max_schema.vehicles (
      id UUID DEFAULT gen_random_uuid() PRIMARY KEY,
      type movr.max_schema.vtype,
      creation_time TIMESTAMPTZ,
      available BOOL,
      last_location STRING
  );

Note that, in addition to the PRIMARY KEY constraint, the id column has a DEFAULT constraint. This constraint sets a default value for the column to a value generated by the gen_random_uuid() function, following UUID best practices. The values generated by this function are guaranteed to be unique and well-distributed across the cluster. We discuss the DEFAULT constraint more below.

Add additional constraints

In addition to the PRIMARY KEY constraint, CockroachDB supports a number of other column-level constraints, including CHECK, DEFAULT, FOREIGN KEY, UNIQUE, and NOT NULL. Using constraints can simplify table queries, improve query performance, and ensure that data remains semantically valid.

To constrain a single column, add a constraint keyword to the column's definition, as shown in the single-column PRIMARY KEY example above. To constrain more than one column, add the entire constraint's definition after the list of columns in the CREATE TABLE statement, also shown in the composite PRIMARY KEY example above.

For guidance and examples for each constraint, see the sections below.

Tip:

For detailed reference documentation for each supported constraint, see the constraint's syntax page.

Populate with default values

To set default values on columns, use the DEFAULT constraint. Default values enable you to write queries without the need to specify values for every column.

When combined with supported SQL functions, default values can save resources in your application's persistence layer by offloading computation onto CockroachDB. For example, rather than using an application library to generate unique UUID values, you can set a default value to be an automatically-generated UUID value with the gen_random_uuid() SQL function. Similarly, you could use a default value to populate a TIMESTAMP column with the current time of day, using the now() function.

For example, in the vehicles table definition in max_init.sql, you added a DEFAULT gen_random_uuid() clause to the id column definition. This set the default value to a generated UUID value. Now, add a default value to the creation_time column:

icon/buttons/copy
CREATE TABLE movr.max_schema.vehicles (
      id UUID DEFAULT gen_random_uuid() PRIMARY KEY,
      type movr.max_schema.vtype,
      creation_time TIMESTAMPTZ DEFAULT now(),
      available BOOL,
      last_location STRING
  );

When a row is inserted into the vehicles table, CockroachDB generates a random default value for the vehicle id, and uses the current timestamp for the vehicle's creation_time. Rows inserted into the vehicles table do not need to include an explicit value for id or creation_time.

Tip:

For detailed reference documentation on the DEFAULT constraint, including additional examples, see the DEFAULT syntax page.

Reference other tables

To reference values in another table, use a FOREIGN KEY constraint. FOREIGN KEY constraints enforce referential integrity, which means that a column can only refer to an existing column.

For example, suppose you want to add a new table that contains data about the rides that MovR users are taking on vehicles. This table should probably include information about the location and duration of the ride, as well as information about the vehicle used for the ride.

In max_init.sql, under the CREATE TABLE statement for vehicles, add a definition for a rides table, with a foreign key dependency on the vehicles table. To define a foreign key constraint, use the REFERENCES keyword:

icon/buttons/copy
CREATE TABLE movr.max_schema.rides (
      id UUID DEFAULT gen_random_uuid() PRIMARY KEY,
      vehicle_id UUID REFERENCES movr.max_schema.vehicles(id),
      start_address STRING,
      end_address STRING,
      start_time TIMESTAMPTZ DEFAULT now(),
      end_time TIMESTAMPTZ
  );

The vehicle_id column will be identical to the id column in the vehicles table. Any queries that insert a vehicle_id that does not exist in the id column of the vehicles table will return an error.

Foreign keys cannot reference tables in a different database. They can, however reference tables in a different schema.

Suppose that you want to introduce promotional codes for users on the MovR platform, but you want the user promo code data to be managed by the abbey user that you created in Create a Database.

Create an empty .sql initialization file for abbey.

icon/buttons/copy
$ touch abbey_init.sql

This file will initialize the objects in the abbey_schema user-defined schema that you created in Create a Schema.

In a text editor, open abbey_init.sql, and add a CREATE TABLE statement for a table called user_promo_codes:

icon/buttons/copy
CREATE TABLE movr.abbey_schema.user_promo_codes (
    code STRING,
    user_email STRING REFERENCES movr.max_schema.users(email),
    valid BOOL,
    CONSTRAINT "primary" PRIMARY KEY (code, user_email)
  );

This new table references the email column of the users table in max_schema. Because the user_promo_codes table depends on the users table, you'll need to execute max_init.sql before abbey_init.sql.

Note:

Foreign key dependencies can significantly impact query performance, as queries involving tables with foreign keys, or tables referenced by foreign keys, require CockroachDB to check two separate tables. We recommend using them sparingly.

Tip:

For detailed reference documentation on the FOREIGN KEY constraint, including additional examples, see the FOREIGN KEY syntax page.

Prevent duplicates

To prevent duplicate values in a column, use the UNIQUE constraint.

For example, suppose that you want to ensure that the email addresses of all users are different, to prevent users from registering for two accounts with the same email address. Add a UNIQUE constraint to the email column of the users table:

icon/buttons/copy
CREATE TABLE movr.max_schema.users (
    first_name STRING,
    last_name STRING,
    email STRING UNIQUE,
    CONSTRAINT "primary" PRIMARY KEY (first_name, last_name)
);

Attempting to insert email values that already exist in the users table will return an error.

Note:

When you add a UNIQUE constraint to a column, CockroachDB creates a secondary index on that column, to help speed up checks on a column value's uniqueness.

Also note that the UNIQUE constraint is implied by the PRIMARY KEY constraint.

Tip:

For detailed reference documentation on the UNIQUE constraint, including additional examples, see the UNIQUE syntax page.

Prevent NULL values

To prevent NULL values in a column, use the NOT NULL constraint. If you specify a NOT NULL constraint, all queries against the table with that constraint must specify a value for that column, or have a default value specified with a DEFAULT constraint.

For example, if you require all users of the MovR platform to have an email on file, you can add a NOT NULL constraint to the email column of the users table:

icon/buttons/copy
CREATE TABLE movr.max_schema.users (
    first_name STRING,
    last_name STRING,
    email STRING UNIQUE NOT NULL,
    CONSTRAINT "primary" PRIMARY KEY (first_name, last_name)
);
Note:

Note that the NOT NULL constraint is implied by the PRIMARY KEY constraint.

Tip:

For detailed reference documentation on the NOT NULL constraint, including additional examples, see the NOT NULL syntax page.

Execute CREATE TABLE statements

After you have defined CREATE TABLE statements for your tables, you can execute the statements.

CREATE TABLE execution best practices

Here are some general best practices to follow when executing CREATE TABLE statements:

Execute the example CREATE TABLE statements

After following the examples provided in the sections above, the max_init.sql file should look similar to the following:

icon/buttons/copy
CREATE TABLE movr.max_schema.users (
    first_name STRING,
    last_name STRING,
    email STRING UNIQUE NOT NULL,
    CONSTRAINT "primary" PRIMARY KEY (first_name, last_name)
  );

CREATE TYPE movr.max_schema.vtype AS ENUM ('bike', 'scooter', 'skateboard');

CREATE TABLE movr.max_schema.vehicles (
      id UUID DEFAULT gen_random_uuid() PRIMARY KEY,
      type movr.max_schema.vtype,
      creation_time TIMESTAMPTZ DEFAULT now(),
      available BOOL,
      last_location STRING
  );

CREATE TABLE movr.max_schema.rides (
      id UUID DEFAULT gen_random_uuid() PRIMARY KEY,
      vehicle_id UUID REFERENCES movr.max_schema.vehicles(id),
      start_address STRING,
      end_address STRING,
      start_time TIMESTAMPTZ DEFAULT now(),
      end_time TIMESTAMPTZ
  );

To execute the statements in the max_init.sql file, run the following command:

icon/buttons/copy
$ cockroach sql \
--certs-dir={certs-directory} \
--user=max \
--database=movr \
-f max_init.sql

The SQL client will execute any statements in max_init.sql, with movr as the database and max as the user. max is now the owner of all objects created by the statements in the max_init.sql file.

After the statements have been executed, you can see the tables in the CockroachDB SQL shell.

Open the SQL shell to your cluster, with movr as the database and max as the user:

icon/buttons/copy
$ cockroach sql \
--certs-dir={certs-directory} \
--user=max \
--database=movr

To view the tables in the max_schema user-defined schema, issue a SHOW TABLES statement:

icon/buttons/copy
> SHOW TABLES FROM max_schema;
  schema_name | table_name | type  | owner | estimated_row_count
--------------+------------+-------+-------+----------------------
  max_schema  | rides      | table | max   |                   0
  max_schema  | users      | table | max   |                   0
  max_schema  | vehicles   | table | max   |                   0
(3 rows)

To see the individual CREATE TABLE statements for each table, use a SHOW CREATE statement. For example, to see the vehicles CREATE TABLE statement:

icon/buttons/copy
> SHOW CREATE TABLE movr.max_schema.vehicles;
         table_name        |                        create_statement
---------------------------+------------------------------------------------------------------
  movr.max_schema.vehicles | CREATE TABLE max_schema.vehicles (
                           |     id UUID NOT NULL DEFAULT gen_random_uuid(),
                           |     type max_schema.vtype NULL,
                           |     creation_time TIMESTAMPTZ NULL DEFAULT now():::TIMESTAMPTZ,
                           |     available BOOL NULL,
                           |     last_location STRING NULL,
                           |     CONSTRAINT vehicles_pkey PRIMARY KEY (id ASC)
                           | )
(1 row)

After following the examples provided in the sections above, the abbey_init.sql file should look similar to the following:

icon/buttons/copy
CREATE TABLE movr.abbey_schema.user_promo_codes (
    code STRING,
    user_email STRING REFERENCES movr.max_schema.users(email),
    valid BOOL,
    CONSTRAINT "primary" PRIMARY KEY (code, user_email)
  );

To execute the statement in the abbey_init.sql file, run the following command:

icon/buttons/copy
$ cockroach sql \
--certs-dir={certs-directory} \
--user=abbey \
--database=movr \
-f abbey_init.sql

After the statements have been executed, you can see the table in the CockroachDB SQL shell.

Open the SQL shell to your cluster, with movr as the database and abbey as the user, and view the table:

icon/buttons/copy
$ cockroach sql \
--certs-dir={certs-directory} \
--user=abbey \
--database=movr
icon/buttons/copy
> SHOW TABLES FROM abbey_schema;
  schema_name  |    table_name    | type  | owner | estimated_row_count
---------------+------------------+-------+-------+----------------------
  abbey_schema | user_promo_codes | table | abbey |                   0
(1 row
icon/buttons/copy
> SHOW CREATE TABLE movr.abbey_schema.user_promo_codes;
              table_name             |                                              create_statement
-------------------------------------+--------------------------------------------------------------------------------------------------------------
  movr.abbey_schema.user_promo_codes | CREATE TABLE abbey_schema.user_promo_codes (
                                     |     code STRING NOT NULL,
                                     |     user_email STRING NOT NULL,
                                     |     valid BOOL NULL,
                                     |     CONSTRAINT "primary" PRIMARY KEY (code ASC, user_email ASC),
                                     |     CONSTRAINT user_promo_codes_user_email_fkey FOREIGN KEY (user_email) REFERENCES max_schema.users(email)
                                     | )
(1 row)

Note that none of the tables that you have created thus far have secondary indexes. For guidance on adding secondary indexes, see Secondary Indexes.

What's next?

You might also be interested in the following pages:


Yes No
On this page

Yes No