AWS’s elastic database cloud thickens

Madan Sheina, Lead Analyst, Software – Information Management

Amazon Web Services (AWS) announced a series of updates to its expanding database services portfolio at its AWS re:Invent conference in Las Vegas, reinforcing its ability to run multiple database workloads with elastic scale in the AWS cloud. While AWS offers a comprehensive menu of best-of-breed cloud services, Ovum would like to see a more integrated platform approach. AWS has disrupted the price barriers for large-scale data processing in the cloud. But cost will become less of a differentiator as customers look for richer and more orchestrated database services to fully support their enterprise data-processing needs.

Momentum and growth continue for AWS database services

Since the launch of Amazon S3 in 2006 as a simple online storage service for software developers, AWS has broadened and deepened its cloud services to cover virtually any type of workload in the cloud – compute, networking, storage, database, application services, and management.

Ovum is seeing more business analytics and optimization-driven use cases being driven by AWS’s portfolio of managed database services (RDS, DynamoDB, Redshift, and ElastiCache). Licensed separately, each is designed for different data-processing workloads in the AWS cloud – supporting relational, NoSQL, analytic (data warehousing), and cached database models.

AWS does not disclose revenue for database services, making it difficult to quantify actual customer adoption. But the company claims strong uptake, notably for Amazon Redshift, which is now the fastest-growing service in terms of revenue in AWS’s history. AWS still has ample opportunity to cross-/upsell its database services within its vast Amazon EC2 installed base, in particular to IT organizations that are comfortable with cloud models, notably Salesforce and NetSuite, which lack robust cloud data-warehousing and BI capabilities.

Advancing its database services remains an investment priority

Acknowledging customer requests, AWS has worked hard to advance its various database services, with a laundry list of improvements.

Amazon RDS, launched in 2009, is AWS’s workhorse managed service for running relational databases (initially Oracle, Microsoft SQL Server, and MySQL) at data loads of up to 3TB. AWS has enriched RDS with cross-region snapshots, replication services, and improved SLAs. Responding to customer demand, AWS is adding a new database engine for running open source Postgres and cross-regional, read-replica snapshots for MySQL.

Amazon DynamoDB, a newer service launched in 2012, is a fully managed service for NoSQL databases. AWS has fleshed the service out with features like parallel scanning, geospatial data indexing, and more granular access controls. AWS has dropped pricing for DynamoDB, which was designed initially as a simple key value store, and is expanding its query capabilities with a new global secondary index, allowing developers to author applications that query against any indices in the table.

Amazon Redshift, the headlining data-warehousing service announcement at last year’s AWS re:Invent, has been upgraded with 40 enhancements in areas such as remote loading, workload memory management, table distribution, JSON and improved CSV support, AWS Data Pipeline integration, and database auditing and logging. AWS has expanded regional support for Redshift in Europe and Asia-Pacific; enhanced concurrency, performance, and provisioning; and improved encryption key management via integration with AWS CloudHSM (Hardware Security Module) and on-premises HSMs. AWS has also extended cross-regional backup support to Redshift customers.

Amazon ElastiCache, a managed caching service introduced in 2011, is often used in conjunction with RDS in web application scenarios. AWS has supplemented it with new cache node types and auto-discovery Redis support for caching of NoSQL data, and has dropped the price, while also offering a free usage tier.

Database services will benefit from other new AWS initiatives

Several other announcements made at AWS re:Invent are also highly relevant to AWS’s database services. The most notable is Amazon Kinesis, a brand-new managed service for processing large volumes of streaming data in realtime, using the AWS cloud to absorb the workload. AWS has built connectors for Kinesis-buffered data to Redshift and DynamoDB, paving the way for customers to ingest, store, and analyze in-flight data. This is a big addition for AWS, and potentially positions Redshift as a realtime data-warehousing platform that will compete against traditionally on-premises solutions of this kind.

AWS also announced its new Amazon CloudTrail, an event-monitoring service for AWS resources for governance and compliance that tracks and logs API calls stored on Amazon S3. Redshift acts as both a provider of CloudTrail event feeds, but more importantly, as a consumer of CloudTrail logs for analytics. The relative infancy of data security in the cloud makes CloudTrail an important development, particularly for securing customers that are still hesitant to push sensitive BI data beyond corporate firewalls.

Maturity, integration, and orchestration needed

Amazon continue to invest in AWS, arguably Amazon’s fastest growth business opportunity, with over 200 updates in 2013 alone and an expanding global infrastructure that now covers 9 regions, 26 availability zones, and 42 edge locations. To keep up with demand, AWS claims to be adding enough new server capacity every day to support Amazon’s global infrastructure when it was a $7bn business!

However, AWS database services are still relatively new and the sheer number of enhancements and additions made to them, particularly around Redshift, suggest they are still maturing. This is a necessary evolution for AWS database services to evolve from being a departmentally focused solution to garnering more full-fledged enterprise-grade database and analytic implementations. That seems to be happening; since Redshift’s launch a year ago, AWS has witnessed a significant increased movement from tentative proof-of-concept deployments to live production systems.

However, Ovum would also like to see these services offered in a more integrated platform-based approach. Reflecting AWS’s “one-size-doesn’t-fit-all” philosophy, these services are offered as individual building blocks, for developing data-driven transactional and analytic applications. Ovum would like to see a more orchestrated approach across AWS’s various cloud services, which rely heavily on third-party tooling from Informatica, SnapLogic, and SyncSort for data transfer and integration. If AWS gets it right, database services like Redshift could well become a lead service that encourages customers to move to the cloud, especially for sensitive BI data.

Related Stories

Leave a comment


This will only be used to quickly provide signup information and will not allow us to post to your account or appear on your timeline.