In this interview Billy Newport, a distinguished engineer and architect at IBM, talks with Floyd Marinescu on topics of virtualization, eXtreme Transaction Processing (XTP) and WebSphere Virtual Enterprise (formerly XD). Billy defines the terms virtualization and hypervisor and goes on to discuss different types of virtualization in an application architecture stack like hardware, operating system, hypervisor, JVM, application and data virtualization.
Billy talked about application virtualization and how WebSphere XD supports features like flow control to dynamically routing the requests based on response times:
So application virtualization, virtualizes a set of hardware resources, but does not take the application knowledge into account. So it knows about may be all the requests flow through a particular kind of router like we have one with XD called On Demand Router and it collects metrics on how long particular requests are taking and then the customer can configure SLA's or Service Level Agreements that say the check out button once you check out, you should have a fast response times of no more than a half second and then what application virtualization lets us do is to measure the individual request going in and because we know which JVMs are hosting those requests, the first level of kind of a reaction we can do is simple flow control where we can start delaying sending servlet requests for less important URI's so we do 2 checkouts for every 1 browser or 3 checkouts so we can give checkout more CPU but then that's really good because obviously we can do that with very little impact and it is something we can do almost in a millisecond.
Speaking about data virtualization offered by products like ObjectGrid and Coherence and Terracotta and GigaSspaces, he said:
..it's broken up into 2 niches right now. There's the majority niche which I characterize as network attach caches which are groups of machines that are organized collectively as one logical cache so they can have as much data as fits in all the boxes in the cache and that everybody that looks at the cache sees one version of every record because it's only stored in one place and they have quality of services like fault tolerance to replication.
And about the second type of data virtualization which is XTP:
XTP is kind of a style of architecture that is becoming more apparent now, it tends to be intrusive whereas the network attached caching, we try to do it in an unobtrusive manner, so you don't have to change your applications ideally, you plug caching into existing and architected interception points and things like that. XTP is different because they were trying to design software that scales in an unlimited manner and that means that you can't just take an application and suddenly make it scale to a million transactions a second but it just doesn't work like that, so you have to have an application designed with a certain paradigm in mind and the XTP paradigm is like a partition data model.
Speaking about WebSphere ObjectGrid product Billy said that it fits into the network attach cache and XTP type scenarios and it's an embeddable platform which makes it easy to deploy in any application server infrastructure. He also said ObjectGrid has agents that can be used for application monitoring, health management and diagnostic information capturing requirements to measure things like java heap and memory utilization.
Here is the interview which is 51 minutes in length.