Newsletters   Subscriptions  Forums  Store   Career  Media Kit  About Us  Contact  Search   Home 
tlb
Volume 2, Number 39 -- October 18, 2005

VMware Boosts VM Scalability with ESX Server 3


by Timothy Prickett Morgan


Virtualization software maker VMware this week gave a sneak preview of the future updates it will be making for its flagship ESX Server virtual machine partitioning software for servers. The company's VirtualCenter partitioning management software was also previewed, and the VMware subsidiary of disk array maker EMC is also boasting about the widespread support for its software among independent software vendors. VMware also announced a set of services to help customers cope with server consolidation and capacity planning.

The ESX Server software is a hypervisor layer that rides on bare metal servers and provides processor instruction set virtualization, memory virtualization, and I/O virtualization for isolated virtual machine partitions. Companies can load most operating systems that are supported on 32-bit X86 processors and 64-bit X64 processors into these virtual machine partitions, and then, using VirtualCenter, dynamically allocate resources for those partitions. VMware also sells another version of virtualization for servers, called GSX Server (and which is at the Release 3 level as well) that carves out virtual machine partitions from within a Windows or Linux operating system running on a 32-bit X86 or 64-bit X64 platform. Because GSX Server runs inside an instance of Windows or Linux, all of the partitions inside GSX Server can be knocked out of those operating systems fail. ESX Server runs on VMware's own hypervisor abstraction layer and provides complete software isolation between the operating systems that run on a server. GSX Server is aimed at development environments and ESX Server is aimed at production environments, although for some production environments GSX Server is probably suitable.

With ESX Server 2.5, which was launched in December 2004, VMware delivered virtual machine partitioning that could have a VM span one or two processors (the latter requiring the VirtualSMP feature) and carve each processor up into a number of individual partitions that was limited mainly by the amount of main memory and CPU cycles an operating system needed to run a workload. With ESX Server 2.5, companies could dedicate as much as 3.6 GB of physical main memory to each partition. ESX Server 2.5 was also equipped with VMotion extensions, which allow the workload from one VM running on one physical machine to be teleported--while it is running--to another physical machine in the network with ESX Server also running on that box. ESX Server 2.5 also included support for storage area networks, which meant that the storage capacity dedicated to VMs could be physically out on a Fibre Channel-attached SAN, which itself can be used to partition storage and provide virtualized access to data.

With ESX Server 3, due early next year, enterprise customers are going to get a high-end hypervisor that can scale further. According to Brian Byun, vice president of product marketing at VMware, ESX Server 3 will support up to four processor cores in a single VM, double that of ESX Server 2.5. (Those cores can be in four separate processors or in two dual-core processors--VMware's software doesn't care how the cores get in the box, it just counts them and uses them.) Main memory allocation for each partition is also extended significantly, with up to 16 GB of main memory per VM partition. What this means is that ESX Server can now be used to carve up a very big server with as many as 16 or 32 X86/X64 processors into machines suitable to be run as application servers for large industries or back-end database servers for midrange shops. VMware's ESX Server 2.5 supports the following operating systems in its partitions: Microsoft Windows XP Pro, Windows NT 4.0, Windows 2000 Server, Windows Server 2003; Red Hat 7.X, 8.0, 9.0 and Enterprise Linux 2.1 and 3; Novell SUSE Linux 8.2, 9.0, and 9.1 as well as Linux Enterprise Server 8 and 9 and NetWare 5.1, 6.0, and 6.5; and FreeBSD 4.9. VMware did not say what operating systems will be supported on ESX Server 3, but Red Hat's Enterprise Linux 4 and Fedora Core 4 seem certain, Novell's Open Enterprise Server (the hybrid NetWare-Linux operating system) as well as SUSE Linux 10.0 seem likely, and even FreeBSD 5.4 or 6.0 could happen.


While ESX Server 2.5 added support for SAN-attached storage, ESX Server 3 will add in support for network-attached storage (NAS) arrays that link to virtualized servers over normal IP networks as well as arrays that link to servers using the iSCSI protocol. The server consolidations that VMware engendered have a rippling effect on storage. "ESX Server actually drives a lot of first-time SAN customers because once they start consolidating and virtualizing their servers, they see they need to consolidate and virtualize their storage," explains Byun. Perhaps the addition of NAS and iSCSI support with ESX Server 3 will also drive these technologies--particularly iSCSI. No one wants to install Fibre Channel SCSI if they can just use Gigabit Ethernet to link their storage and servers together over iSCSI; that is why iSCSI was created, after all. Not everyone needs 4 Gb/sec Fibre Channel bandwidth for their storage.

By the way, what ESX Server 3 does not yet include is support for the Virtualization Technology (VT) electronics for supporting instruction set virtualization due in Intel's Xeon line of chips next year; nor will it support the similar "Pacifica" instruction set virtualization electronics due in the next rev of the Opteron chips from AMD. VMware will, of course, support these new features of future chips some time after they become available. But exactly when remains unclear.

Byun says that the previews of ESX Server 3 and VirtualCenter 2 are just to help companies better plan their future installations, and given the substantial improvements VMware is adding, a lot of companies might wait and do their installations with ESX Server 3 a few months from now instead of ESX Server 2.5 now. ESX Server 3 is in a limited beta testing program right now and will go into a broader beta program by the end of the year. VMware is targeting the first quarter of 2006 for the general availability of ESX Server 3 and VirtualCenter 2. Pricing information for the future software was not pre-announced, but VMware recently said that even though it counts cores as processors in terms of how its VMs work, in terms of pricing it will treat a single socket as the main unit of per-processor pricing.

The main enhancements with VirtualCenter 2 are something VMware is calling distributed availability services. With this feature, VMware is baking high availability into ESX Server, VirtualCenter, VMotion, and VirtualSMP and if an application load fails inside of a VM, VirtualCenter will automatically detect that failure and restart the applications inside another partition on another suitable machine in the network. Because disk access for VMs is already virtualized, this will eliminate the need to high availability disk clustering for a lot of customers (provided they mirror their data at the hardware level). Customers are not just concerned about server uptime, of course, but how well applications are running and meeting service-level agreements (SLAs). So VirtualCenter 2 can monitor application performance inside the VMs and move VMs around the server network to meet targeted SLAs as workloads change over the course of a day, month, or year; this feature is called distributed resource scheduling. VMware says a single instance of VirtualCenter 2 can control hundreds of physical servers and thousands of virtual ones.

GSX Server is not being updated this time around and is already at the Release 3 level (which was introduced in February 2004).

To help customers cope with the complexities of server virtualization and consolidation, and to boost its revenues, VMware is also announcing this week assessment services that use a hosted capacity planning tool created by VMware to monitor a prospective customer's applications as they are running and to suggest possible configurations for a consolidated, virtualized platform that can handle the same workload. VMware and its partners are using this tool to help streamline the process for customers. The quicker you can size up a customer and get it virtualized, the quicker you can move on to the next customer (or the more you can do in parallel, however you want to look at it).

Finally, VMware is bragging that over 60 operating system, database, middleware and application software suppliers have tested and certified their software to work inside VMware's virtual machines. There are, of course, many more applications that work inside of VMs--including homegrown code.

Sponsored By
LINUX NETWORX

Clusterworx® Whitepaper

High performance Linux clusters can consist of hundreds or thousands of individual components. Knowing the status of each CPU, memory, disk, fan, and other components is critical to ensure the system is running safely and effectively.

Likewise, managing the software components of a cluster can be difficult and time consuming for even the most seasoned administrator. Making sure each host's software stack is up to date and operating efficiently can consume much of an administrator's time. Reducing this time frees up system administrators to perform other tasks.

Though Linux clusters are robust and designed to provide good uptime, occasionally conditions lead to critical, unplanned downtime. Unnecessary downtime of a production cluster can delay a product's time to market or hinder critical research.

    Since most organizations can't afford these delays, it's important that a Linux cluster comes with a robust cluster monitoring tool that:
  • Provides essential monitoring data to make sure the system is operational.
  • Eliminates repetitive installation and configuration tasks to reduce
          periods of downtime.
  • Provides powerful features, but doesn't compromise on usability.
  • Automates problem discovery and recovery on would-be critical events.

This paper discusses the features and functions of Clusterworx® 3.2. It details how Clusterworx® provides the necessary power and flexibility to monitor over 120 system components from a single point of control. The paper also discusses how Clusterworx® reduces the time and resources spent administering the system by improving software maintenance procedures and automating repetitive tasks.

High Performance Monitoring

Each cluster node has its own processor, memory, disk, and network that need to be independently monitored. This means individual cluster systems can consist of hundreds or thousands of different components. The ability to monitor the status and performance of all system components in real time is critical to understanding the health of a system and to ensure it's running as efficiently as possible.

Because so many system components need to be monitored, one of the challenges of cluster management is to efficiently collect data and display system health status in an understandable format. For example, let's say a cluster system has 100 nodes and is running at 97 percent usage. It's very important to know whether 100 nodes are running at 97 percent usage or whether 97 nodes are running at 100 percent usage while three nodes are down.

Clusterworx® provides real-time analysis of over 120 essential system metrics from each node. Data is displayed in easy-to-read graphs, thumbnails, and value tables. Clusterworx® collects data from groups of nodes to spot anomalies, then drills down to single node view to investigate problems. This allows users to determine exactly what the problem is before taking corrective action.

Clusterworx® also tracks the power and health state of each node and displays its status using visual markers in a node tree view throughout the user interface. Power status shows whether the node is on, off, provisioning, or in an unknown state. The health state tracks informational or warning messages and critical errors. Health state messages are displayed in a message queue on the interface.

Clusterworx®'s comprehensive monitoring and easy-to-read charts and graphs allow users to quickly asses the state of each node and the overall system at a glance - while providing the necessary information to make informed decisions about the cluster system.

To read the rest of this whitepaper, please visit www.linuxnetworx.com
Editor: Timothy Prickett Morgan
Contributing Editors: Dan Burger, Joe Hertvik, Kevin Vandever,
Shannon O'Donnell, Victor Rozek, Hesh Wiener, Alex Woodie
Publisher and Advertising Director: Jenny Thomas
Advertising Sales Representative: Kim Reed
Contact the Editors: To contact anyone on the IT Jungle Team
Go to our contacts page and send us a message.

THIS ISSUE
SPONSORED BY:

Arkeia
Roaring Penguin
Linux Networx
Egenera
OpenLogic


The Linux Beacon

BACK ISSUES

TABLE OF
CONTENTS
Three Mandriva 2006 Linux Editions Come to Market

IBM, Novell Offer Chassis-Level Linux Pricing on Blades

VMware Boosts VM Scalability with ESX Server 3

Mad Dog 21/21: New Moth

But Wait, There's More


The Four Hundred
The "P" Word

IBM Gives Rebates and Trade Ins to Push the i5 520 in Q4

Why i for the Casino Industry?

Stop Arguing About Cars and Start Managing Fleets

The Windows Observer
Microsoft to Adapt Server Licensing for Virtualized Environments

Intel Begins Dual-Core Xeon Server Chip Rollout

Patch Tuesday Yields Nine Patches, Three That Are Critical

Microsoft Unveils New Security Tools and Security Vendor Consortium

The Unix Guardian
Big Iron Still Costs Big Bucks

Intel Begins Dual-Core Xeon Server Chip Rollout

Server Makers Are Ready and Sorta Eager for Dual-Core Xeons

Ich Bin Ein Entrepreneur


Copyright © 1996-2008 Guild Companies, Inc. All Rights Reserved.
Guild Companies, Inc. (formerly Midrange Server), 50 Park Terrace East, Suite 8F, New York, NY 10034
Privacy Statement