AWS Redshift · Rate Limits

Aws Redshift Rate Limits

Amazon Redshift enforces account- and cluster-level quotas per AWS Region. Database throughput is bounded by node count, RPU sizing, and WLM configuration. The Redshift Data API enforces fixed per-second throttles per AWS account per Region (e.g. ExecuteStatement 30 TPS, DescribeStatement 100 TPS, ListTables 3 TPS). Provisioned cluster connection ceilings are 2,000 for RA3 nodes and 500–2,000 for DC2 by node size.

Aws Redshift Rate Limits is the machine-readable rate-limit profile for AWS Redshift on the APIs.io network, conforming to the API Commons Rate Limits specification.

It captures 23 rate-limit definitions, measuring nodes, clusters, snapshots, query_slots, and connections.

The profile also includes 5 backoff/retry policies defined and response codes documented for throttled and serviceUnavailable.

Tagged areas include Data Warehouse, Analytics, Rate Limiting, and Throttling.

23 Limits Throttle: 429
Data WarehouseAnalyticsRate LimitingThrottling

Limits

Nodes per region (provisioned) account/region
nodes
200
Across all clusters. Increasable.
RA3 nodes in a cluster cluster
nodes
128
Increasable.
DC2 nodes in a cluster cluster
nodes
128
Increasable.
Concurrency scaling clusters account/region
clusters
10
Increasable.
Reserved nodes account/region
nodes
200
Increasable.
Snapshots account/region
snapshots
700
Increasable.
WLM concurrency (manual queues) cluster
query_slots
50
Across all user-defined manual WLM queues. Not increasable.
Connections per RA3 cluster cluster
connections
2000
Connections per DC2 cluster cluster
connections
2000
dc2.large = 500; dc2.8xlarge = 2,000.
Serverless base RPU per account account
RPUs
3200
Quota is the greater of 3,200 RPUs or 1.5x your max aggregate base RPUs over the prior six months. Increasable.
Serverless workgroups account/region
workgroups
25
Increasable.
Serverless namespaces account/region
namespaces
25
Increasable.
Data API ExecuteStatement account/region
requests_per_second · second
30
Not increasable.
Data API BatchExecuteStatement account/region
requests_per_second · second
20
Not increasable.
Data API GetStatementResult account/region
requests_per_second · second
20
Not increasable.
Data API DescribeStatement account/region
requests_per_second · second
100
Not increasable.
Data API CancelStatement account/region
requests_per_second · second
3
Not increasable.
Data API DescribeTable / ListDatabases / ListSchemas / ListStatements / ListTables account/region
requests_per_second · second
3
Each individual List/Describe API in this group is capped at 3 TPS. Not increasable.
Idle session timeout (Provisioned) cluster
hours
4
User setting can override.
Idle session timeout (Serverless) workgroup
hours
1
Idle transaction timeout cluster
hours
6
Serverless query timeout workgroup
seconds
86399
24 hours maximum running query.
Spectrum single COPY row size query
MB
4

Policies

WLM (Workload Management)
Throttling and queueing of concurrent SQL is managed via WLM (manual or auto). Tune queues, query priorities, and short-query acceleration to balance interactive vs ETL.
Concurrency Scaling
When the main cluster's queues fill, Redshift can transparently dispatch read queries to concurrency-scaling clusters. First 1 hour/day free per main cluster.
Backoff for Data API
AWS SDKs retry Data API throttles automatically; raw clients should implement exponential backoff with jitter on ThrottlingException / 429.
Service Quotas Increases
Most cluster-level quotas (nodes, RPUs, snapshots, concurrency-scaling clusters) are adjustable via Service Quotas. WLM concurrency, connection caps, and Data API per-API TPS are not adjustable.
Datashare Discipline
Cross-account / cross-region datashares avoid cross-cluster ETL and are preferable to transferring data via S3 in many analytics use cases.

Sources