Elastic Data Warehouse | Elasticity at Scale

Instant and Online

Handle sudden spikes in data queries or continuous growth in data volume without performance degradation.

Performant and Predictable

Leverage engine observability features to optimize performance and scale your workloads efficiently, while reducing costs.

Cost-Effective

Scale and de-allocate resources online—only pay for what you use, eliminating cost overruns due to compute overprovisioning.

Multidimensional Elasticity

Independently scale your compute resources across multiple dimensions, such as up/down, out/in, and concurrency scaling with consistent performance predictability and zero service interruptions.

CREATE ENGINE IF NOT EXISTS MyEngine
WITH TYPE = S NODES = 1 CLUSTERS = 1;

Scale-Up/Down

Adjust compute node sizes to handle varying workloads. Scale up to increase node sizes or scale down to reduce costs.

ALTER ENGINE MyEngine SET TYPE = L;

Scale-Out/In

Add or remove compute nodes to adjust performance and cost. Scale out to increase nodes, scale in to reduce them.

ALTER ENGINE MyEngine SET NODES = 2;

Concurrency Scaling

Additional clusters are added to handle more simultaneous queries, ensuring high query throughput without performance degradation.

ALTER ENGINE MyEngine SET CLUSTERS = 2;

Learn more

Granular Scaling

Scale one node at a time to adjust compute resources incrementally, ensuring an ideal price-performance ratio tailored to your evolving workload requirements.

Right size your engines

If your workload requires more than one L node, there's no need to double capacity to two L nodes. Instead, right-size your configuration with five S type nodes, costing 37.5% less than two L nodes. Since one L node has resources equivalent to four S nodes, you can right-size your configuration with five S nodes.

ALTER ENGINE MyEngine SET TYPE
= S NODES = 5;

Engine Observability

Optimize performance with real-time insights into engine metrics. Leverage built-in views, accessible through information_schema, to correctly size your engines, ensuring both optimal performance and cost efficiency.

SELECT * FROM information_schema.engine_metrics_history

SELECT * FROM information_schema.engine_running_queries

SELECT * FROM information_schema.engine_query_history

Online Scaling and Upgrades

Experience consistent performance and zero downtime with advanced upgrades and instant scaling, optimizing your infrastructure to handle any workload without service interruptions.

Dynamic Scaling

Instantly adjust resources to match workload requirements, avoiding over-provisioning.

Online Transition

System automatically directs traffic to new configurations while gracefully phasing out the old ones.

Zero Downtime

Automatic upgrades without service interruptions.

Transparent Process

Built-in validation of workload performance before engine upgrades

Automatic Start and Stop

Eliminate idle time costs and pay only for the compute you actively use with auto start/stop functionality. Go online instantly while scaling resources to match your needs.

Instant

Stop using compute resources when idle, and quickly restarting to handle varying workloads.

Configurable

Fine-tune engine settings and optimize costs to match your workload requirements.

Automatically start engine

ALTER ENGINE my_prod_engine WITH
AUTO_START = TRUE;

Automatically stop engine

ALTER ENGINE my_prod_engine WITH
AUTO_STOP = 20;

Instant

Stop using compute resources when idle, and quickly restarting to handle varying workloads.

Configurable

Fine-tune engine settings and optimize costs to match your workload requirements.

Automatically start engine

ALTER ENGINE my_prod_engine WITH
AUTO_START = TRUE;

Automatically stop engine

ALTER ENGINE my_prod_engine WITH
AUTO_STOP = 20;

Elastic Scaling for Dynamic Workloads

Multidimensional Elasticity

Granular Scaling

Engine Observability

Online Scaling and Upgrades

Automatic Start and Stop

Tune in to the 'Data Engineering Show' to see how the fastest growing tech companies handle their data challenges
Real talk, no fluff.

Elastic Scaling for Dynamic Workloads

Multidimensional Elasticity

Granular Scaling

Engine Observability

Online Scaling and Upgrades

Automatic Start and Stop

Tune in to the 'Data Engineering Show' to see how the fastest growing tech companies handle their data challengesReal talk, no fluff.

Tune in to the 'Data Engineering Show' to see how the fastest growing tech companies handle their data challenges
Real talk, no fluff.