IBM code unfetters virtual workloads

IBM has cracked the problem of moving virtual machines across different servers, without using shared storage

By Joab Jackson, IDG News Service |  Storage Add a new comment

Some of the first fruits of a European Union-funded project led by IBM are making their way into the field of cloud computing, in the form of a virtual machine migration technology.

The technology, sprouting from the Reservoir (Resources and Services Virtualization without Barriers) program, offers a way to move a live, virtualized workload from one server to another, without the need for the two locations to share the same storage space.

"We at IBM Research view this as a huge leap forward because the new open source code enables sharing across domains where shared networked storage and hardware components are not practical," wrote Dr. Yaron Wolfsthal, an IBM research senior manager of system technologies and services who led the development, in a blog post announcing the code.

"This is a technology for live service migration which obviates the need for shared storage between the source and target physical machines [which] was previously required by all known migration mechanisms," Wolfsthal said in a subsequent e-mail interview.

Of course, VMware has offered the ability to move a virtual machine from one server to another for a while now, by way of software called VMotion. But this approach has its limits: It requires both the source and the destination hosts to share the same storage device, which limits the range of migration. "To the best of our understanding, vSphere does not support live migration without shared storage," Wolfsthal said.

In fact VMware has been trying to solve the long-distance problem. Last year it said it was working with Cisco Systems on a proof-of-concept demonstration that it planned to show off at a Cisco event in San Francisco. "This, of course, is a non-trivial thing to do," VMware said at the time.

Others have commented on the single-storage-location limitation as well. "This means that inside a single data center, one can only move a [virtual machine] across a relatively small number of physical machines. Not exactly what the marketing guys would have you believe," wrote Alex Benik, a principal at analyst firm Battery Ventures, in a research note.

In contrast, the Reservoir technology involves copying the workload over to the new location -- while keeping it running -- using a delicately choreographed combination of push and pull replication mechanisms. "The movement is not restricted and virtual machines can be now moved from anywhere to anywhere," Wolfsthal explained. Both locations will need to have the code running in their hypervisors in order to make this migration work.

Thus far, the capability has been integrated into two open-source virtualization programs. One is KVM (Kernel-based Virtualization), a hypervisor technology that has been incorporated into the Linux kernel and is the cornerstone of Red Hat's virtualization strategy. It has also been inserted into the Libvirt virtualization toolkit, which supports both the Citrix Xen hypervisor and the VMware hypervisors.

Wolfsthal declined to talk about what, if any, plans IBM has for commercializing the technology.

Reservoir is a three-year European Union-funded program to develop new cloud computing, virtualization and Web 2.0 technologies.

Joab Jackson covers enterprise software and general technology breaking news for The IDG News Service. Follow Joab on Twitter at @Joab_Jackson. Joab's e-mail address is Joab_Jackson@idg.com

    Add a comment

    Post a comment using one of these accounts
    Or join now
    At least 6 characters

    Note: Comment will appear soon after you have activated your account.
    Obscene/spam comments will be removed and accounts suspended.
    The information you submit is subject to our Privacy Policy and Terms of Service.

    ITworld LIVE

    StorageWhite Papers & Webcasts

    White Paper

    ESG ~ HP StoreOnce: the Next Wave of Data Deduplication

    Leveraging deduplication in backup environments yields significant advantages. The cost savings in reducing disk capacity requirements change the economics of disk-based backup. For some organizations, it allows disk-based backup-and, importantly, recovery-to be extended to additional workloads in the environment. For others, deduplication makes it possible to introduce disk-based backup where it may not have been feasible before.

    White Paper

    Evaluator Group: Storage Federation - IT Without Limits (Analysis of HP Peer Motion with Storage Federation)

    As the role of IT increases within organizations, the need to move data when and where it is needed is critical to support emerging business requirements. This has become increasingly difficult due to the huge growth of data volumes. This white paper sponsored by HP + Intel evaluates a solution that aims to enable the movement of data without physical limitations. Read now and see how this could enable agility and efficiency.

    White Paper

    HP Converged Storage Sets the Stage for the Next Era of Computing

    Enterprise storage has undergone many changes in recent years - with converged storage and infrastructure 2.0 paving the way for reduced IT infrastructure costs and greater performance. This report discusses the latest trends that are setting the stage for the next era of computing. Learn about the new infrastructure and storage trends that are changing the way business storage works today.

    White Paper

    AppAssure vs Acronis

    In this study of data protection for environments with virtual and physical servers running Windows, openBench Labs tested AppAssure Backup and Replication software v 4.7 and Acronis Backup & Recovery 11. Both solutions utilize block-based technology to unify data protection operations.

    White Paper

    Guaranteeing 100% Backup Recovery

    The single biggest challenge for IT personnel involved in the data protection process is making sure that their backups are recoverable every time. Management and users won't remember the ninety-nine successful recoveries but they will always remember the one failure.

    See more White Papers | Webcasts

    Ask a question

    Ask a Question