Adaptive Computing Moves into Big Data Workflows with Moab 6.0
Software as a service
platform provider Adaptive Computing on Feb. 14 released a new version of its
frontline Moab Adaptive Computing Suite, which has been redesigned to handle
so-called "big data" private-cloud system deployments-meaning
petabyte-level and larger storage and workloads.
The new Moab software package is optimized to satisfy growing enterprise demand for deploying IAAS (infrastructure as a service), PAAS (platform as a service) or SAAS (software as a service) from a centralized and intelligent data center infrastructure, Peter ffoulkes, vice president of marketing for Adaptive Computing, told eWEEK.
"The thing that's really important about Moab is its intelligent decision-making ability," ffoulkes said. "This comes into play once you get past what we call the 'Cloud 1.0 level' [usually meaning a standard portal plus a few basic features, like chargeback and agility features] and move into the big data space.
"In the large systems, you've got idle virtual machines sitting there burning up resources with nobody using them, servers with applications that have been shut down, or that were 80 percent loaded earlier in the day and are now 20 percent loaded at the end of the day ... all sorts of adjustments that need to be made," ffoulkes said. "This all happens as workflows come and go. They expand and contract in their demand levels. That adaptiveness is a critical element to the correct functionality of a large enterprise private cloud, and that is what Moab brings."
Moab Adaptive Computing Suite is powered by Moab, the company's core unified intelligent automation technology.
New features in Moab v6.0 include the automatic tuning to manage a mix of application workloads; the automatic initiation of live migration of virtual machine workloads to meet service needs; the rapid automation of resource allocation, provisioning and de-provisioning hardware resources via policies; the provision of billing or "showback" of resource-use costs; and the aggregation of data for rich context to automate initial resource allocation decisions.
"Moab was originated for managing this kind of work in the HPC [high-performance computing] world. Now we have extended that into the capabilities of Moab 6.0," ffoulkes said. "In addition to the batch work we've always been good at, we now support transactional workflows by linking applications sets together, so that when requests come in, say from a Web farm, we can manage that flow all the way backward and forward."
By ensuring that applications and infrastructure are automatically deployed properly the first time, ffoulkes said, service failures are avoided, resource use is optimized and bottom-line costs are reduced.