Forgot your password?
typodupeerror
Operating Systems Software Virtualization

Inside VMware's 'Virtual Datacenter OS' 121

Posted by Soulskill
from the closer-look dept.
snydeq writes "Neil McAllister cuts through VMware's marketing hype to examine the potential impact of VMware's newly pronounced 'virtual datacenter OS' — which the company has touted as the death knell for the traditional OS. Literally an operating system for the virtual datacenter, VDC OS is an umbrella concept to build services and APIs that make it easier to provision and allocate resources for apps in an abstract way. Under the system, McAllister writes, apps are reduced to 'application workloads' tailored through vApp, a tool that will allow developers to 'encapsulate the entire app infrastructure in a single bundle — servers and all.' The concept could help solve the current bugbear of programming, parallel processing, McAllister concludes, assuming VMware succeeds."
This discussion has been archived. No new comments can be posted.

Inside VMware's 'Virtual Datacenter OS'

Comments Filter:
  • by tji (74570) on Saturday September 20, 2008 @11:29AM (#25085115)

    They are not replacing Linux. You still run your what you want on Linux, but do you run everything on ONE Linux box? If yes, you're not a good candidate for a Datacenter OS. If you run many servers, then there is almost definitely room for efficiency in that structure.

    Rather than dedicating the full bare hardware to your App, you deploy as a VM in your Virtual Datacenter ( mini cloud ). The DCOS takes care of managing the resources, things like:

    - Moving your server VM from compute node to compute node to automatically balance load and optimize performance,
    - Move VMs to work around failures, allow hardware upgrades, etc. without downtime.
    - Expand capacity by dropping another compute node into the cloud (the big difference between the old mainframe world and the new DCOS. This scales easily with cheap powerful nodes)
    - Move the machine images around your storage infrastructure, to allow for management, maintenance, upgrade, expansion, etc.
    - Provide recovery and even fault tolerance of hardware. Servers can automatically move and re-start on hardware failure; or they can even run in lockstep to maintain full operation through a node failure.

    This is VMware's big lead (and big need to leverage, as the revenue from the hypervisor layer dries up). They provide the management layer that enables all the above, and they keep improving it. From a central GUI, I can manage all my VMs, manage the compute resources as a cluster.

  • by Anonymous Coward on Saturday September 20, 2008 @11:42AM (#25085229)

    I have to disagree with AC that "vmware is the only solution", 6 months ago we evaluated both vmware (which we had been using in dev and test for years) and the Citrix Xen product and decided to go for Xen for our production systems based upon performance we saw (yes yes YMMV) cost, and the open nature of the API. The problem was finding a strong partner/integrator to help us swing our server estate from physical to virtual in the time allotted.

    So far the systems have been solid, and required only a couple of noncritical updates/patches to maintain. I agree that its not about the hypervisor (any more), with Hyper-V being practically free ($40-bucks?), but I would disagree that VMWare is the only game in town.

    We were prepared to question the accepted wisdom in search of performance and savings, and it worked out well for us (so far). I can't comment on the other versions of Xen from Sun, Oracle, and Co, but we found Citrix Xen and a hardware virtualization appliance [360is.com] a solid, manageable system.

  • Re:done already (Score:3, Informative)

    by TheRaven64 (641858) on Saturday September 20, 2008 @12:22PM (#25085485) Journal
    Samsung did some work with Xen (not sure if it's published yet) on 'partial migration' where a group of independent machines appeared to be a single SMP (or, ideally, AMP) machine from the perspective of the OS and pages were synchronised between different nodes using a cache coherency protocol as required. Marathon have a Xen version which synchronises two VM instances in remote locations allowing either to act as a transparent fail-over for the other if one set of hardware goes down. Moving it down to the storage layer doesn't sound like it provides anything more than existing SAN solutions which can be deployed on most VMs already.

    Oh, and Simon's a bright guy, but he works more on the marketing side of things. If you want technical comments, go to Ian or Keir.

  • by ArsonSmith (13997) on Saturday September 20, 2008 @01:06PM (#25085829) Journal

    Not exactly. VDC-OS does actually replace Windows and/or Linux. Think of it as a Linux kernel, and instead of InitV startup your app starts up. You don't maintain users and directories or storage or even log into a shell. The OS is reduced to just enough to run 1 application and only 1 application.

    This OS/App bundle is created with a basic config file and is then started just like you'd start a virtual machine on an ESX server or server cluster. ESX can then handle the migration and resources of all the ESX servers in the cluster and your App will move between them to the one that will service it best.

    I'm sure there will be a debug environtment that may included a more advanced shell like interface to run traces and such, but a stable app shouldn't need any of this. (By stable I mean well tested and developed to one specific environment of which this will provide. Think cartrage console game vs. PC game)

  • by Ritchie70 (860516) on Saturday September 20, 2008 @01:18PM (#25085949) Journal

    I think it is, actually.

    We've got some VMware guys at my job doing a proof of concept for us. (I work for one of those big companies where people hear the name and that cha-ching noise happens in their head.)

    Each VM has its own MAC address, and the virtualization layer includes a network switch. So long as the switch knows where to send the packets, and the other end of the TCP connection is willing to tolerate a few moments of silence while the VM moves, it should work.

  • by image77 (304432) on Saturday September 20, 2008 @01:23PM (#25085987)

    Yep - the "cutover" happens faster that the TCP timeout window. The connection stays alive, and even if a packet is lost it simply gets resent when the ACK goes missing.

  • by Ralish (775196) <ralish@gma[ ]com ['il.' in gap]> on Saturday September 20, 2008 @01:47PM (#25086173)

    To clear up the acronym soup a little:

    HA = High Availability
    Technology that aims to ensure (high) availability of virtual machines across a virtualised cluster through intelligent monitoring of VM's and cluster resources.
    http://www.vmware.com/products/vi/vc/ha.html [vmware.com]

    DR = Distributed Resource Scheduler (I assume that's what parent meant)
    Provides much more advanced and fine-grained control of the available resources in a virtualised cluster.
    http://www.vmware.com/products/vi/vc/drs.html [vmware.com]

    BC = Consolidated Backup (guessing)
    Backup technology for virtualised clusters, providing backup features specific to virtualisation scenarios that conventional backup products don't traditionally offer, although, BC can integrate with them to an extent.
    http://www.vmware.com/products/vi/consolidated_backup.html [vmware.com]

  • by Ralish (775196) <ralish@gma[ ]com ['il.' in gap]> on Saturday September 20, 2008 @01:53PM (#25086211)

    Yes, VMware provides a technology for its datacenter level products called VMotion that does exactly that, moving VM's between physical virtualised servers in a cluster while preserving all active networking connections.

    I don't know the specifics of how it works and manages that feat, but I have seen tech demos that show it in action. I watched one a while ago published by Dell showing a VMotion task in progress, I'm sure you can find it on the web somewhere with some digging around.

    Regardless, it does work and has been available as part of its enterprise products for some time.

    If you want to know more:
    http://www.vmware.com/products/vi/vc/vmotion.html [vmware.com]

  • (Open) Solaris (Score:3, Informative)

    by d3xt3r (527989) on Saturday September 20, 2008 @02:09PM (#25086315)
    Solaris 10 and Open Solaris have the concept of zones and containers. The computer runs a single Solaris instance but can run isolated process trees in zones which share common libraries but can be updated for dependencies independently. The containers concept (in conjunction with zones) allows a fair share scheduler to guarantee a service level for each allocated zone (CPU/memory sharing, etc). IMHO, must better than Virtuozzo, VMware and Xen.
  • by mattmarlowe (694498) on Saturday September 20, 2008 @04:48PM (#25087419) Homepage

    You apparently missed the the announcement from Cisco that they've released their own virtual switch with enterprise features to replace the limited capabilities in VMware's. And, yes, vmware will fully support it and it will be plug and play compatible. Furthermore, on a cluster of ESX hosts, you can have multiple Cisco supervisor appliances running for HA/management, while a Cisco switch configuration/etc is shared across all nodes and ports being logically linked to each vm, regardless of where vm is located and even during vmotion.

    Cisco details at: http://www.cisco.com/en/US/products/ps9902/index.html [cisco.com]

  • by Anonymous Coward on Sunday September 21, 2008 @10:41AM (#25092965)

    Not even close. http://vmware.com/products/thinapp/ [vmware.com]

Luck, that's when preparation and opportunity meet. -- P.E. Trudeau

Working...