Huh? Doesn't IBM Marketing do their homework before they do their spin?
""In the current Hadoop distro, it is one job at a time," Hertzler tells El Reg. "You need to add distributed cluster logic to manage multiple MapReduce jobs at the same time on the same cluster." Or, use multiple Hadoop clusters, as Yahoo! does. "But Symphony is already a distributed workload manager and knows how to distribute data and work around a cluster.""
I guess IBM doesn't know about the Fair Scheduler.
The only feature Hadoop scheduling doesn't offer is the ability to suspend a job so that it can be completed at a later time.
Of course here's the problem... only a portion of IBM's 'Blue Stack' customers are going to be willing to go with IBM's idea of Grid computing. Too many companies have either committed to Hadoop or have committed to using Hadoop in part of a PoC. (Proof of Concept).
I had thought IBM learned their lesson from their earlier mistakes when getting in to this space.
Apparently not.