If your company is using Amazon Redshift for data warehousing and analytics in the cloud, you may be confused by AWS Redshift pricing options. Redshift allows for flexibility and scaling to fit your company’s data needs, from on-demand, hourly rates to discounted rates through reserved instances. Choosing the right pricing plan will help your company save money, as will taking advantage of built-in tools and savvy management of Redshift operations.
Redshift is the most robust data warehouse available, but it can be an expensive proposition, making management and optimization extremely critical. Therefore it is important to understand the different pricing options available to get the most value out of your dbase services on AWS.
How is Redshift Pricing Determined?
Amazon’s Redshift pricing is based on the node type used and the physical location of the stored data. Costs vary greatly by region, though the storage capacity and computing power remains constant, regardless of location. Three types of Redshift nodes are available — RA3, Dense Compute (DC) and Dense Storage (DS) — based on how your data is stored. If you expect your data to grow over time, RA3 nodes will provide the best price and performance. DC nodes are optimized for quicker, more efficient queries. The more expensive DS nodes are recommended for larger stores of data above 500GB. You can add nodes to improve query performance and boost data storage.
What Different Costs are Involved with Redshift?
Amazon Redshift On-Demand pricing lets you pay hourly, based on node usage. Redshift Spectrum, for directly running SQL queries against data in your S3 data lake, is priced at $5.00 per terabyte. With Concurrency Scaling pricing, each cluster of nodes earns up to an hour of free concurrency scaling credits every 24-hour period. Thousands of concurrent users and queries enjoy fast performance, and you only pay the on-demand, per-second rate for usage beyond the free credits. Managed Storage pricing lets you pay an hourly rate for data stored in RA3 clusters, regardless of how many compute nodes are provisioned.
How Does Using On-Demand vs. Reserved Instances Affect Pricing?
With Amazon Redshift on-demand pricing, you pay for storage capacity at an hourly rate, with no upfront cost or commitment. Your rate is calculated based on the type and number of nodes in your data warehouse cluster. You don’t have to plan for or purchase instances before you’re certain of what you need, such as during development or testing. Reserved instances, costing substantially less, are ideal for steady workloads. You can commit to a 1-year or 3-year term, with no upfront payment, make a partial payment or pay in full upfront.
What are the Different Redshift Node Types?
In the Amazon Redshift architecture, there are three node types: RA3, DC2 and DS2. You can choose the right node type according to the performance required, amount of data to be processed and your plans for growth. RA3 nodes with managed storage let you scale your data warehouse storage capacity without adding and paying for additional nodes. DC2 nodes provide data warehouses that are compute-intensive — local high performance SSD storage is included. For companies using DS2 nodes with hard disk drive storage, upgrading to RA3 nodes will provide twice as much storage and improved performance at the same price.
How Can You Reduce Redshift Costs?
Use Reserved Instances
One way to reduce spending is to commit to a set number of nodes for 1- or 3-year term with Amazon’s reserved instances pricing model. The discount will vary based on the upfront payment you can make. The most cost effective option is to make a partial upfront payment. While the discount for a full upfront payment is the highest, the savings differential from partial upfront is not significant, therefore you can preserve part of the initial capital outlay for other investments. Depending on your location and the node type and size, the 1-year plan can save as much as 42% over on-demand pricing. With a 3-year plan, the savings can go up to 75%.
However, Reserved Instances do not offer the same flexibility for Amazon Redshift services as they do for Amazon EC2 instances. Unlike compute instances, unused RDS instances cannot be sold in the AWS Marketplace. Therefore all reservations need to be applied to actual dbase usage before their term expires. Otherwise savings from reservations will be reversed.
The below table illustrates the savings available from Reserved Instances per TB of data. Please note that, for RA3, data stored in managed storage is billed separately based on actual data stored in the RA3 node types; effective price per TB per year is calculated for only the compute node costs.
Choose the Right Cluster Type and Size
Choosing the right cluster type, and the proper size, can help a company achieve substantial savings in their AWS spend. This is especially important if you wish to take advantage of savings through Reserved Instances.
Optimize Queries with Well-Structured Tables
Your Redshift cluster performance highly depends on the compute level, so it’s especially important to finely tune your queries. This will also have a profound effect on your Redshift monthly cost.
Pause and Resume Services
The pause and resume feature in the Redshift console lets you suspend billing when your Redshift cluster is not being used for consecutive hours. It can also be automated to match your usage schedule. The pause and resume feature is simplified, so you no longer have to do a backup, terminate the cluster and then restore it from a snapshot. Instead, when compute is suspended, the underlying data and data structures remain intact.
Scale Capacity Based on Load
This involves constant monitoring of your storage and compute usage so that you always have the resources that are necessary for peak performance without paying for more than you need. This can be trickier than you may realize, as it means you must be always analyzing your data processes, which is a time-consuming activity.
How Parquantix Can Help You Get the Most from your AWS Dbase Services, at the Best Price
Parquantix provides companies with deployments on Amazon Web Services with an automated plug-in solution that monitors and manages your resources in real time. Our solution is tailored to your company’s unique requirements. It’s proprietary AI-driven algorithm adapts with every new AWS technology and pricing update to optimize the performance and value of your database services.
Are you ready to start optimizing the cost of your dbase applications? Contact us to schedule a 30-minute consultation.
