virtualization.info Editors

As most of you probably know ESX Server requires supported SCSI disks for running VMFS, the VMware proprietary file system where virtual machines are stored.
This slows down the spread of ESX Server in home testing environments and the upcoming ESX Server 3.0 won’t help much, introducing support for iSCSI and NAS but not for IDE/SATA.

But since more than 1 year there is known trick to workaround the limitation: using some special SATA/SATA2 controllers which use SCSI drivers.

Richard Garsthagen, a VMware employee, published a funny story about how trying to get ESX working with a SATA controller can actually be much more expensive than buying a SCSI subsystem.

By the way he also confirmed that the SATA2 MegaRAID SATA 300-8x controller works great for ESX 2.x and upcoming ESX 3.0 (at least on the actual beta)!

Thincomputing.net compiled a list of settings to modify for avoid time synchronization in a virtual machine.

You probably would have this working (and I strongly recommend to do so in a production environment or wherever you’re working with security tools like firewalls, IDS, SEM and so on) but sometimes for testing purposes could be useful stop the synchronization.

Read it here.

Ben Armstrong details those factors you should consider when choosing a virtual disk for your virtual machines in Virtual Server.
As you’ll read IDE is limited in disks quantity and disk size but is faster and better supported than SCSI.

Read it here.

Quoting from the SWsoft official annoncement:

IT professionals can now move virtual servers between physical servers without end-user disruption or the need for costly storage capacity with the release today of Virtuozzo for Linux 3.0 from SWsoft — a recognized leader in server virtualization and automation software.

The “zero downtime migration” capability is one of several major new features in the latest version of Virtuozzo for Linux – which also adds new template management, streamlined resource management and expanded operating system support to greatly enhance performance and automation.

SWsoft also introduced Virtuozzo for Windows 3.5.1 – which adds a “physical-to-virtual” tool that enables easy migration from a dedicated physical server to a virtual server, performance increases and updated platform support.
…

What is New in Virtuozzo for Linux 3.0
This latest version features improved performance and automation capabilities, including the following features and benefits:

• Zero Downtime Migration, which captures the state of an existing virtual server and its contents and migrates it to a new physical server without any interruption in service or availability. The function executes between any two servers on a network, so the capability works for any server and any application. SWsoft is the first to deliver this capability without additional requirements, such as a storage area network (SAN).
• EZtemplates make it easy for almost instant (generally 20-30 seconds) provisioning of a virtual server, so it is easy to set up and run applications on a virtual server, rather than a full server.
• Streamlined resource management that requires just a few settings to create and manage a virtual server.
• Support for distributions based on the Linux 2.6 kernel including Red Hat Enterprise Server 4.0, Fedora Core 4.0 and CentOS 4.0 and backward compatibility for 2.4 distributions.

What is New in Virtuozzo for Windows 3.5.1
This latest version includes the following features and benefits:

• New graphical VZP2V (Physical-to-Virtual) tool for administrators to easily move from a dedicated physical server to a Virtuozzo virtual server. This feature helps determine what resources a server and application require to ensure that the virtual server delivers the same performance.
• Performance increases for accessing memory and handling read/write operations, as well as making it even faster to create and migrate a virtual server.
• Support for Microsoft Windows Server 2003 R2 and Microsoft Small Business Server 2003 so that the software can be run in a virtual server.

What is New in Virtuozzo Management Tools
Extensive upgrades to Virtuozzo’s suite of management tools include Virtual Server Provisioning Workflow, expanded Virtuozzo Control Center, Firewall Configuration enhancements, Multi-Virtual Server Commands, Desktop Configuration and several additional features.

Pricing and Availability
Virtuozzo for Linux 3.0, Virtuozzo for Windows 3.5.1 and Virtuozzo Management Tools are now available worldwide from SWsoft or any of its partners. Pricing for Virtuozzo starts at $1000/CPU with management tools at $1000/seat for unlimited servers.

Quoting from the Parallels official announcement email:

Following the successful release of Parallels Workstation 2.0, the world’s first hypervisor powered desktop virtualization solution, Parallels Inc., today announced the availability of Parallels Workstation 2.1 beta. The new version of Parallels’ powerful, user-friendly, cost-effective desktop virtualization solution boasts a number of important improvements, including:

• Stronger virtual machine performance, stability and isolation
• USB support enables users to access a wide range of popular USB devices directly in virtual machines
• “Suspend and Resume” functionality that lets users instantly suspend a virtual machine’s state and resume it at a later time without having to completely shut down and start up a guest operating system
• Support for Physical Address Extension (PAE) allows Workstation 2.1 to run on PAE-enabled primary OSes with up to 4GB of RAM
• Wireless networking support in Windows frees users from network cables. Now, users can connect virtual machines to the internet via any WiFi connection
• Stronger support for Intel VT technology and a more complex virtual processor significantly improve performance of a wide range of applications running in virtual machines
• An improved, redesigned Parallels Tools package for Windows that provides useful add-on features that let each user customize their Parallels Workstation experience. New tools include guest/host time synchronization and complete support for normal video modes in Windows 95, 98 and Me
• Primary and guest OS support for SuSE Linux 10, and guest OS support for Sun Solaris

Apply for the beta program here.

Quoting from the DataCore official announcement:

DataCore Software today announced at Server Centric Consulting’s Geek Day, an all-day technology showcase taking place today in St. Louis, the general availability of powerful Windows data migration capabilities that enable users to move and migrate live data disks over a SAN to other Windows systems across their organization. The new capabilities, launched today as SANmotion, offer a simplified and centralized approach to perform Windows volume administration by supporting new ‘ready to use’ and pre-existing Windows formatted data volumes. Apart from eliminating work and saving IT administrators time, this advanced feature allows Windows disk data to be accessed, re-purposed or moved rapidly across the SAN or over Ethernet connections (over iSCSI).
…
The SANmotion technology is now available in all new SANmelody™ 2.0 software releases, DataCore’s disk server storage virtualization and management solution. “This product empowers users to move their data and disks around the SAN to wherever they need them,” said George Teixeira, president and CEO, DataCore Software. “The analogy to this would be how VMware, through VMotion, enables the movement of a PC from one place to another. SANmotion delivers disk resources with live data over the SAN. This eliminates the burden that a Windows administrator would have of formatting and setting up disk volumes for each and every individual computer. Instead, one admin does it centrally saving the disruption, time and motion to do the same thing on each separately located desktop or server.”…

Thank to Steven Bink for the news.

This is what Keith Adams, a VMware engineer, said on his personal blog (so be aware that this is not an official statement):

…
I’ve written a lot here about VT, Intel’s recently shipped CPU virtualization hardware. I’m pretty intimately familiar with VT’s gory guts, as well as those of AMD’s Pacifica, after spending a good chunk of my career at VMware extending our VMM to support them. (Sorry to disappoint, AMD/Intel fanboys: the two specifications are pretty much exactly the same, with different instructions and in-memory layouts, etc.).
…
I was also faintly aware that IBM had done some work to accelerate virtualization “back in the day.” But, I was utterly shocked at the familiarity of this paper (Osisek, Jackson, Gum, in IBM Systems Journal, March, 1991). It describes interpretive execution, which is IBM’s name for the S/390 virtualization acceleration hardware. What’s fascinating is that “interpretive execution” so closely resembles Pacifica, and in turn VT, that you can mechanically translate among them.

What VT calls “non-root mode”, and Pacifica calls “guest mode”, is known as “interpretive execution” (which, by the way, joins a long list of nuttily technical-sounding, yet completely non-descriptive terms that I associate with IBM; it’s right up there with “translation lookaside buffer”). VT’s “vmlaunch” instruction is Pacifica’s “vmrun” is s/390’s Germanic-flavored “sie”; Intel’s “VMCS” is AMD’s “VMCB” is IBM’s “state description” (another hilarious IBM-ism)…

Read the whole post at source.

Quoting from the AMD official annoucement:

AMD today announced the broad availability of its I/O virtualization technology specification. Demonstrating continued leadership in evolving the industry-standard x86 processor architecture, AMD is making this technology available through royalty-free licenses that are designed to encourage its widespread adoption by hardware and software developers. By defining a set of technologies that can be directly built into a computer’s I/O bridges, AMD’s I/O virtualization technology is designed to extend the benefits of CPU-assisted virtualization by addressing the performance bottlenecks and security issues that can be encountered when virtualizing I/O devices in x86-based servers, desktops, and notebook computers.
…
AMD’s I/O virtualization technology is expected to be supported by all AMD processors in mid-2006, and is also anticipated to be implemented in chipsets and core logic designed for AMD64-based platforms in 2006. The AMD I/O virtualization technology specification download, together with technology overviews and guidance to software developers who are designing virtualization solutions for 64-bit technology, can be found at http://developer.amd.com.

In November I already pointed out AMD official slides reporting part of the virtualization strategy roadmap.
On this AnandTech article you can find a much more detailed roadmap providing raw dates, socket details (the reported M2 socket is now known as AM2), and processors codenames.

Meanwhile AMD collected a nice amount of Pacifica material on the dedicated website:

• AMD Pacifica Virtualization Technology overview
• AMD: WebBench Virtualization Performance Study
• Virtualization with AMD & VMware paper

I also remind you that Pacifica specifications are available here.

Quoting from eWeek:

Users with servers powered by Intel’s “Paxville” Xeon MP processor can now activate the chip’s virtualization capabilities.

Speaking at a conference here on virtualization, Diane Bryant, vice president and general manager of Intel’s Server Platforms Group, said that users can now contact systems makers Dell, Hewlett-Packard and IBM to help them upgrade the BIOS on the chip to enable the activation of the Intel Virtualization Technology.

The technology was in the chip when it started shipping last year, but it was disabled until now.

“You can now enable virtualization technology in your servers and start running it in your testing and development environments,” Bryant told the conference attendees.

She also noted support from Red Hat and Novell’s SuSE Linux unit for the open-source Xen virtualization technology, and announced virtualization software vendor VMware’s release Feb. 6 of its free VMware Server product…

Read the whole article at source.