After “day 0”, partner day, VMworld Europe opened its doors for the general audience today.
vSphere is the official name of the new VMware platform, but the VMware marketing department has not applied their new naming convention vPrefix to all products & initiatives yet. Some VMware partners are already doing the same, such as Intel with vPro. How long will it take for Vmworld to be renamed vWorld. Below are the six building blocks of vSphere:
- vCompute (hardware assisted virtualization and extended live migration compatibility):
- vStorage (storage management and replication)
- vNetwork (for network management, look for Cisco here…
- Security (where VMsafe innovates on firewalls, anti-virus, intrusion detection/prevention and compliance)
- Scalability (dynamic resource sizing)
- Availability (data protection and clustering)
Here’s a rush rundown of the keynote by VMware CEO Paul Maritz.
The theme of the conference is a continuation of VMworld 2008 in Las Vegas: “Virtualy anything is possible.” The unofficial theme seems to be the same as in Las Vegas as well: “cloud, cloud, cloud”.
Even though the next release of ESX is just around the corner, we don’t expect major announcements on that scale today.
Update: The stream of today’s keynote is now available.
Maurizio Carli, General Manager VMware EMEA
09.11 Maurizio Carli, General manager EMEA takes the stage.
Last year 4500 people attended VMworld Europe. Even in the current economic climate, 4700 people showed up this week.
9:23 Paul once again takes the blame for the massive proliferation of x86 servers in the datacenters. (Paul Maritz is a 14-year Microsoft veteran)
9:25 In the early 2000s, hypervisors introduced the concept of consolidation. Maritz points out that this is the point where most of VMware’s competition is now. VMware is now talking about “cooperating hypervisors”, and, of course: the cloud.
9:26 The VMware vision is that the Cloud will be built on industry-standard building blocks, starting with the “internal cloud”, based on the Virtual Datacenter OS.
VMware likes to call this VDC-OS the “software mainframe”.
9:27 When internal environments are “converted” to this VDC_OS, it becomes easier to take the encapsulated workloads and migrate them to external, federated cloud providers, in a non-disruptive way.
9:28 VMware knows that hardware and a hypervisor aren’t enough, but that security policies, quality of service and management are just as important.
9:30 Virtualization is the key to making this happen in an evolutionary way: existing applications can be put in the “Black boxes” virtualization provides.
– vSphere architecture
9:33 The product name for the new generation of VDC-OS products will be: vSphere. No surprises there..
9:39 No new stuff so far… vSphere requires a new management suite, now called the vCenter suite. (As opposed to VirtualCenter).
Apparently some stuff will be demoed later. Curious…
General overview of the VMworld stage
9:44 The second initiative, a logical extension of the VDC-OS, is vCloud, where customers will have the choice to go to an external service provider to get their IT infrastructure. VMware aims to build compatible clouds (based on VDC-OS of course), allowing users to build private clouds, where external and internal IT resources are pooled together and managed as one.
9:45 VMware will work with the formal standards bodies to make sure users aren’t locked in to one vendor’s cloud. There should be a broad ecosystem of clouds, giving users choice to move in and out of clouds as necessary.
9:46 (And again, it seems like Amazon EC2 doesn’t exist, even though with them “cloud” is a reality today, sort of.)
9:47 The first guest comes on stage, Kurt Glazemakers, EMEA CTO of terremark.
9:49 
Terremark CTO Kurt Glazemakers
9:50 
Terremark enterprise cloud
9:51 Pooling resources on a hosting platform gives users the possibility to leverage economies of scale of large environments, providing ample burst capacity if necessary. This reduces provision times. Users don’t have to worry about CAPEX, as the server capacity is treated like a service (OPEX).
9:52 Terremark created a self-service portal allowing users to create VMs as they please, within the limits of their resource pool. Users pay by the GHz of CPU power and GB of memory and disk storage.
9:56 Next guest: Joe Arnold, director of Engineering of Engine Yard, a Ruby on Rails company.
10:00 Engine Yard created a self-service portal to create RoR containers. Pretty short demo. Looks a bit like CohesiveFTs Elastic Server.
10:00 Another guest on stage: Zvi Guterman, CEO IT Structures
10:05 Paul gives some more examples of service providers. Savvis – one of the biggest hosting companies building a giant resource pool for customer VMs. Sungard, providing disaster recovery solutions as a service.
10:06 The third leg of the future VMware stragey is the vClient initiative.
10:08 The management of user workloads should not be done at the device level, but at the user level. The workloads should follow the user wherever he is and whatever device he’s using.
10:09 VMware started as a client-side virtualization company, with “VMware”, now VMware Workstation.
10:09 To allow “offline VDI”, VMware will provide a client-side bare metal hypervisor. (A la Phoenix?)
10:10 This enables the user to checkout his desktop when working on a mobile device, and to check in and work on a thin client when at the office, leveraging central management and intelligent storage (with deduplication, …)
10:13 All the vClient / VMware View stuff announced so far (WAN optimization, thin client optimization, offline VDI, …) should be rolled out completely in 2009.
10:13 No news on the semi-recent mobile hypervisor acquisition so far.
10:14 New announcement: formal partnership with Intel.
Guest on stage: Gregory Bryant, VP Business Client Group at Intel.
10:16 VMware and Intel will work together on a client-side hypervisor.
10:18 The collaboration enables out-of band, centralized management, but gives the user the genuine local desktop experience.
Check back tomorrow for a more in-depth presentation by Stephen Herrod, VMware’s CTO.
Good recap of the conference. Thanks! Regarding your comments from 9:51 to 10:05, there are several other companies providing services (pooling infrastructure, pay-per-use) across the market spectrum. These include JumpBox and click2try (http://www.click2try.com). As far as I know, click2try is the only one currently focused on the Open Source community, trying to evangelize the use of enterprise-level OSS community versions from Alfresco, Pentaho, OrangeHRM, Nagios, and others. Also, the only one that has opted to build the infrastructure using the Open Source version of the Xen hypervisor.