Quotas and limits in Amazon Redshift - Amazon Redshift
Quotas for Amazon Redshift objectsQuotas for Amazon Redshift Serverless objectsQuotas for Amazon Redshift Data APIQuotas for query editor v2 objectsQuotas and limits for Amazon Redshift Spectrum objectsNaming constraints
Services or capabilities described in Amazon Web Services documentation might vary by Region. To see the differences applicable to the China Regions, see Getting Started with Amazon Web Services in China (PDF).

Quotas and limits in Amazon Redshift

Amazon Redshift has quotas that limit the use of several resources in your Amazon account per Amazon Region. There is a default value for each quota and some quotas are adjustable. For adjustable quotas, you can request an increase for your Amazon account in an Amazon Region by submitting an Amazon Redshift Limit Increase Form.

Quotas for Amazon Redshift objects

Amazon Redshift has quotas that limit the use of several object types. There is a default value for each.

Quota name Amazon default value Adjustable Description

Amazon accounts that you can authorize to restore a snapshot per snapshot

20

No

The maximum number of Amazon accounts that you can authorize to restore a snapshot, per snapshot.

Amazon accounts that you can authorize to restore a snapshot per Amazon KMS key

100

No

The maximum number of Amazon accounts that you can authorize to restore a snapshot, per KMS key. That is, if you have 10 snapshots that are encrypted with a single KMS key, then you can authorize 10 Amazon accounts to restore each snapshot, or other combinations that add up to 100 accounts and do not exceed 20 accounts for each snapshot.

Cluster IAM roles for Amazon Redshift to access other Amazon services

501

No

The maximum number of IAM roles that you can associate with a cluster to authorize Amazon Redshift to access other Amazon services for the user that owns the cluster and IAM roles.

1The quota is 10 in the following Amazon Web Services Regions: us-iso-east-1, us-iso-west-1, us-isob-east-1.

Concurrency level (query slots) for all user-defined manual WLM queues

50

No

The maximum query slots for all user-defined queues defined by manual workload management.

Concurrency scaling clusters

10

Yes

The maximum number of concurrency scaling clusters.

DC2 nodes in a cluster

128

Yes

The maximum number of DC2 nodes that you can allocate to a cluster. For more information about node limits for each node type, see Clusters and nodes in Amazon Redshift.

DS2 nodes in a cluster

128

Yes

The maximum number of DS2 nodes that you can allocate to a cluster. For more information about node limits for each node type, see Clusters and nodes in Amazon Redshift.

Event subscriptions

20

Yes

The maximum number of event subscriptions for this account in the current Amazon Region.

Nodes

200

Yes

The maximum number of nodes across all database instances for this account in the current Amazon Region.

Parameter groups

20

No

The maximum number of parameter groups for this account in the current Amazon Region.

RA3 nodes in a cluster

128

Yes

The maximum number of RA3 nodes that you can allocate to a cluster. For more information about node limits for each node type, see Clusters and nodes in Amazon Redshift.

Redshift-managed VPC endpoints connected to a cluster

30

Yes

The maximum number of Redshift-managed VPC endpoints that you can connect to a cluster. For more information about Redshift-managed VPC endpoints, see Working with Redshift-managed VPC endpoints.

Grantees to cluster accessed through a Redshift-managed VPC endpoint

5

Yes

The maximum number of grantees that a cluster owner can authorize to create a Redshift-managed VPC endpoint for a cluster. For more information about Redshift-managed VPC endpoints, see Working with Redshift-managed VPC endpoints.

Redshift-managed VPC endpoints per authorization

5

Yes

The maximum number of Redshift-managed VPC endpoints that you can create per authorization. For more information about Redshift-managed VPC endpoints, see Working with Redshift-managed VPC endpoints.

Reserved nodes

200

Yes

The maximum number of reserved nodes for this account in the current Amazon Region.

Schemas in each database per cluster

9,900

No

The maximum number of schemas that you can create in each database, per cluster. However, pg_temp_* schemas do not count towards this quota.

Security groups

20

Yes

The maximum number of security groups for this account in the current Amazon Region.

Single row size when loading by COPY

4

No

The maximum size (in MB) of a single row when loading by using the COPY command.

Snapshots

20

Yes

The maximum number of user snapshots for this account in the current Amazon Region.

Subnet groups

20

Yes

The maximum number of subnet groups for this account in the current Amazon Region.

Subnets in a subnet group

20

Yes

The maximum number of subnets for a subnet group.

Tables for large cluster node type

9,900

No

The maximum number of tables for the large cluster node type. This limit includes permanent tables, temporary tables, datashare tables, and materialized views. External tables are counted as temporary tables. Temporary tables include user-defined temporary tables and temporary tables created by Amazon Redshift during query processing or system maintenance. Views and system tables aren't included in this limit.

Tables for xlarge cluster node type

9,900

No

The maximum number of tables for the xlarge cluster node type. This limit includes permanent tables, temporary tables, datashare tables, and materialized views. External tables are counted as temporary tables. Temporary tables include user-defined temporary tables and temporary tables created by Amazon Redshift during query processing or system maintenance. Views and system tables aren't included in this limit.

Tables for xlplus cluster node type with a single-node cluster.

9,900

No

The maximum number of tables for the xlplus cluster node type with a single-node cluster. This limit includes permanent tables, temporary tables, datashare tables, and materialized views. External tables are counted as temporary tables. Temporary tables include user-defined temporary tables and temporary tables created by Amazon Redshift during query processing or system maintenance. Views and system tables aren't included in this limit.

Tables for xlplus cluster node type with a multiple-node cluster.

20,000

No

The maximum number of tables for the xlplus cluster node type with a multiple-node cluster. This limit includes permanent tables, temporary tables, datashare tables, and materialized views. External tables are counted as temporary tables. Temporary tables include user-defined temporary tables and temporary tables created by Amazon Redshift during query processing or system maintenance. Views and system tables aren't included in this limit.

Tables for 4xlarge cluster node type

200,000

No

The maximum number of tables for the 4xlarge cluster node type. This limit includes permanent tables, temporary tables, datashare tables, and materialized views. External tables are counted as temporary tables. Temporary tables include user-defined temporary tables and temporary tables created by Amazon Redshift during query processing or system maintenance. Views and system tables aren't included in this limit.

Tables for 8xlarge cluster node type

200,000

No

The maximum number of tables for the 8xlarge cluster node type. This limit includes permanent tables, temporary tables, datashare tables, and materialized views. External tables are counted as temporary tables. Temporary tables include user-defined temporary tables and temporary tables created by Amazon Redshift during query processing or system maintenance. Views and system tables aren't included in this limit.

Tables for 16xlarge cluster node type

200,000

No

The maximum number of tables for the 16xlarge cluster node type. This limit includes permanent tables, temporary tables, datashare tables, and materialized views. External tables are counted as temporary tables. Temporary tables include user-defined temporary tables and temporary tables created by Amazon Redshift during query processing or system maintenance. Views and system tables aren't included in this limit.

Number of databases

60

No

The maximum allowed count of databases in an Amazon Redshift cluster. This excludes databases created from datashares.

Timeout for idle or inactive sessions

4 hours

No

This setting applies to the cluster. For information about setting the idle-session timeout value for a user, see ALTER USER in the Amazon Redshift Database Developer Guide. The user setting takes precedence over the cluster setting.

Timeout for idle transactions

6 hours

No

The maximum period of inactivity for an open transaction before Amazon Redshift ends the session associated with the transaction. This setting takes precedence over any user-defined idle timeout setting. It applies to the cluster.

Stored procedures in a database

10,000

No

The maximum number of stored procedures. See Limits and differences for stored procedure support for more limits.

Maximum number of connections for RA3 nodes

2,000

No

The maximum number of connections to an RA3 cluster. (This applies specifically to the ra3.xlplus, ra3.4xlarge and ra3.16xlarge node types.) The maximum connections allowed varies by node type.

Maximum number of connections for DC2 and DS2 nodes

Varies

No

The maximum number of connections to a dc2.large or ds2.large cluster is 500. The maximum number of collections to a dc2.8xlarge or ds2.8xlarge cluster is 2000.

Number of Amazon Redshift roles in a cluster

1,000

Yes

The maximum number of Amazon Redshift roles that you can create per cluster. For more information about role-based access control (RBAC) roles see Role-based access control (RBAC) in the Amazon Redshift Database Developer Guide

Quotas for Amazon Redshift Serverless objects

Amazon Redshift has quotas that limit the use of several object types in your Amazon Redshift Serverless instance. There is a default value for each.

Quota name Amazon default value Adjustable Description

Number of databases

100

No

The maximum allowed count of databases in an Amazon Redshift Serverless namespace. This excludes databases created from datashares.

Number of schemas

9,900

No

The maximum allowed count of schemas in an Amazon Redshift Serverless instance.

Number of tables

200,000

No

The maximum allowed count of tables in an Amazon Redshift Serverless instance.

Timeout for idle or inactive sessions

1 hour

No

For information about setting the idle-session timeout value for a user, see ALTER USER in the Amazon Redshift Database Developer Guide. The user setting takes precedence.

Timeout for a running query

86,399 seconds (24 hours)

No

The maximum time for a running query before Amazon Redshift ends it.

Timeout for idle transactions

6 hours

No

The maximum period of inactivity for an open transaction before Amazon Redshift Serverless ends the session associated with the transaction. This setting takes precedence over any user-defined idle timeout setting.

Number of maximum connections

2000

No

The maximum number of connections allowed to connect to a workgroup.

Number of workgroups

25

Yes

The number of workgroups supported.

Number of namespaces

25

Yes

The number of namespaces supported.

Number of Amazon Redshift roles in a workgroup

1,000

Yes

The maximum number of Amazon Redshift roles that you can create per workgroup. For more information about role-based access control (RBAC) roles see Role-based access control (RBAC) in the Amazon Redshift Database Developer Guide

For more information about how Amazon Redshift Serverless billing is affected by timeout configuration, see Billing for Amazon Redshift Serverless.

Quotas for Amazon Redshift Data API

Amazon Redshift has quotas that limit the use of the Redshift Data API. There is a default value for each. For more information about Amazon Redshift Data API, see Using the Amazon Redshift Data API.

Quota name Amazon default value Adjustable Description

Transactions per second (TPS) for the BatchExecuteStatement API

20

No

The maximum number of operation requests you can make per second without being throttled.

Transactions per second (TPS) for the CancelStatement API

3

No

The maximum number of operation requests you can make per second without being throttled.

Transactions per second (TPS) for the DescribeStatement API

100

No

The maximum number of operation requests you can make per second without being throttled.

Transactions per second (TPS) for the DescribeTable API

3

No

The maximum number of operation requests you can make per second without being throttled.

Transactions per second (TPS) for the ExecuteStatement API

30

No

The maximum number of operation requests you can make per second without being throttled.

Transactions per second (TPS) for the GetStatementResult API

20

No

The maximum number of operation requests you can make per second without being throttled.

Transactions per second (TPS) for the ListDatabases API

3

No

The maximum number of operation requests you can make per second without being throttled.

Transactions per second (TPS) for the ListSchemas API

3

No

The maximum number of operation requests you can make per second without being throttled.

Transactions per second (TPS) for the ListStatements API

3

No

The maximum number of operation requests you can make per second without being throttled.

Transactions per second (TPS) for the ListTables API

3

No

The maximum number of operation requests you can make per second without being throttled.

Quotas for query editor v2 objects

Amazon Redshift has quotas that limit the use of several object types in your Amazon Redshift query editor v2. There is a default value for each.

Quota name Amazon default value Adjustable Description

Connections

500

Yes

Maximum number of connections that you can create using the query editor v2 in this account in the current Region.

Active principals per account

50

Yes

Maximum number of simultaneous principals who can use query editor v2 in this account in the current Region.

Saved queries

2,500

Yes

Maximum number of saved queries that you can create using the query editor v2 in this account in the current Region.

Query versions

20

Yes

Maximum number of versions per query that you can create using the query editor v2 in this account in the current Region.

Saved charts

500

Yes

Maximum number of saved charts that you can create using the query editor v2 in this account in the current Region.

Rows fetched per query

100,000

No

Maximum number of rows fetched per query by the query editor v2 in this account in the current Region.

Data fetched size per query

5

No

Maximum size, in megabytes, of the data fetched per query by the query editor v2 in this account in the current Region.

Simultaneous socket connections per principal

10

Yes

Maximum number of simultaneous socket connections to query editor v2 that a single principal can establish in the current Region. Evaluate whether to increase this quota if you receive errors that your socket connections are over the limit.

Simultaneous socket connections per account

250

Yes

Maximum number of simultaneous socket connections to query editor v2 that all principals in the account can establish in the current Region. Evaluate whether to increase this quota if you receive errors that your socket connections are over the limit.

Maximum concurrent connections

3

No

Maximum database connections per user (includes isolated sessions). This value can be set from 1–10 by the query editor v2 administrator in Account settings. If you reach the limit set by your administrator, consider using shared sessions instead of isolated sessions when running your SQL. For more information about connections, see Opening query editor v2. For more information about setting the limit, see Changing account settings.

Quotas and limits for Amazon Redshift Spectrum objects

Amazon Redshift Spectrum has the following quotas and limits:

  • The maximum number of databases per Amazon account when using an Amazon Glue Data Catalog. For this value, see Amazon Glue service quotas in the Amazon Web Services General Reference.

  • The maximum number of tables per database when using an Amazon Glue Data Catalog. For this value, see Amazon Glue service quotas in the Amazon Web Services General Reference.

  • The maximum number of partitions per table when using an Amazon Glue Data Catalog. For this value, see Amazon Glue service quotas in the Amazon Web Services General Reference.

  • The maximum number of partitions per Amazon account when using an Amazon Glue Data Catalog. For this value, see Amazon Glue service quotas in the Amazon Web Services General Reference.

  • The maximum number of columns for external tables when using an Amazon Glue Data Catalog, 1,597 when pseudocolumns are enabled, and 1,600 when pseudocolumns aren't enabled.

  • The maximum size of a string value in an ION or JSON file when using an Amazon Glue Data Catalog is 16 KB. The string can be truncated if you reach this limit.

  • You can add a maximum of 100 partitions using a single ALTER TABLE statement.

  • All S3 data must be located in the same Amazon Region as the Amazon Redshift cluster.

  • Timestamps in ION and JSON must use ISO8601 format.

  • External compression of ORC files is not supported.

  • Text, OpenCSV, and Regex SERDEs do not support octal delimiters larger than '\177'.

  • You must specify a predicate on the partition column to avoid reads from all partitions.

    For example, the following predicate filters on the column ship_dtm, but doesn't apply the filter to the partition column ship_yyyymm:

    WHERE ship_dtm > '2018-04-01'.

    To skip unneeded partitions you need to add a predicate WHERE ship_yyyymm = '201804'. This predicate limits read operations to the partition \ship_yyyymm=201804\.

These limits don't apply to an Apache Hive metastore.

Naming constraints

The following table describes naming constraints within Amazon Redshift.

Cluster identifier

  • A cluster identifier must contain only lowercase characters.

  • It must contain 1–63 alphanumeric characters or hyphens.

  • Its first character must be a letter.

  • It cannot end with a hyphen or contain two consecutive hyphens.

  • It must be unique for all clusters within an Amazon account.

Database name

  • A database name must contain 1–64 alphanumeric characters.

  • It must contain only lowercase letters.

  • It cannot be a reserved word. For a list of reserved words, see Reserved words in the Amazon Redshift Database Developer Guide.

Endpoint name of a Redshift-managed VPC endpoint

  • An endpoint name must contain 1–30 characters.

  • Valid characters are A-Z, a-z, 0-9, and hyphen(-).

  • The first character must be a letter.

  • The name can't contain two consecutive hyphens or end with a hyphen.

Admin user name

  • An admin user name must contain only lowercase characters.

  • It must contain 1–128 alphanumeric characters.

  • Its first character must be a letter.

  • It cannot be a reserved word. For a list of reserved words, see Reserved words in the Amazon Redshift Database Developer Guide.

Admin password

  • An admin password must contain 8–64 characters.

  • It must contain at least one uppercase letter.

  • It must contain at least one lowercase letter.

  • It must contain one number.

  • It can use any ASCII characters with ASCII codes 33–126, except ' (single quote), " (double quote), \, /, or @.

Parameter group name

  • A parameter group name must contain 1–255 alphanumeric characters or hyphens.

  • It must contain only lowercase characters.

  • Its first character must be a letter.

  • It can't end with a hyphen or contain two consecutive hyphens.

Cluster security group name

  • A cluster security group name must contain no more than 255 alphanumeric characters or hyphens.

  • It must contain only lowercase characters.

  • It must not be Default.

  • It must be unique for all security groups that are created by your Amazon account.

Subnet group name

  • A subnet group name must contain no more than 255 alphanumeric characters or hyphens.

  • It must contain only lowercase characters.

  • It must not be Default.

  • It must be unique for all subnet groups that are created by your Amazon account.

Cluster snapshot identifier

  • A cluster snapshot identifier must contain no more than 255 alphanumeric characters or hyphens.

  • It must contain only lowercase characters.

  • It must not be Default.

  • It must be unique for all snapshot identifiers that are created by your Amazon account.