Teradata rolled out several new cloud-friendly deployment options for its analytic database at this week's Teradata Partners Conference. Smart, given that earlier this year IDC predicted that global spending on public cloud services will enjoy a 19.4 percent compound annual growth rate (CAGR) from 2015 to 2019.
"Companies need their data and analytic environment to be agile, multi-faceted, and flexible," said Oliver Ratzesberger, executive vice president and chief product officer at Teradata, in a statement. "Teradata Everywhere means that the exact same Teradata Database can be deployed on-premises and in the cloud. ... Companies can concentrate their energy on running analytics to improve their business rather than engineering compatibility between their environments."
The Teradata Everywhere initiative supports multiple deployment environments, including Amazon Web Services (AWS), Microsoft Azure, Teradata Managed Cloud and VMware.
Though Teradata Database was already available on AWS in multiple different geographic regions, customers can now choose a massively parallel processing (MPP) configuration that scales up to 32 nodes. New features for AWS deployments include automatic node failure recovery; backup, restore and querying of data in Amazon Simple Storage Service (S3); and software ecosystem launch templates that enable one-click deployment automation. The company is also expanding the geographies in which it is available; AWS GovCloud, South America and China are now the only regions in which it cannot be deployed.
The same MPP configuration will be available on Microsoft Azure by the end of this year, in all regions except US Gov Iowa, West India and China.
Coming online this quarter is a virtual machine edition, with a preconfigured VM image for VMware environments. The product offers flexible licensing options so companies can scale configurations to match changes in requirements and also employ multiple instances for special purposes such as testing.
Teradata is also beefing up its own IntelliFlex cloud service that it says offers independent scaling of processing power and storage capacity, advanced in-memory computing and reduced downtime for system expansions. Both performance density and memory will increase, with now up to 12 nodes per cabinet and up to a terabyte of memory per node. In addition, all configurations will include SSD for enhanced query responsiveness. Teradata expects all new IntelliFlex features to be available by the end of this year.
To ensure smooth performance and flexibility across the various deployment options, Teradata also announced it is enhancing core features of the Teradata Database.
It is adding stream query re-planning, which modifies query steps during query execution using actual versus estimated results, to its Teradata Database Adaptive Optimizer feature. Also, adaptive platform-specific costing will recognize and adapt to the host platform, enabling Teradata Everywhere to deliver optimized platform-specific performance. According to Teradata, targeted availability is Q2 of 2017.
Finally, a MAPS feature will provide elasticity to expand or shrink a Teradata data warehouse to scale alongside changing business needs. According to the company, MAPS will eliminate data redistribution time following a system expansion by performing table-at-a-time online reconfiguration, selectively redistributing tables at the most convenient time. This feature should also be available in 2017's second quarter.