Trends in virtualization focusing on solutions, not technologies.The technology of virtualization An umbrella term for enhancing a computer's ability to do work. Following are the ways virtualization is used. Hardware Virtualization Partitioning the computer's memory into separate and isolated "virtual machines" simulates multiple machines within one physical computer. is well understood and deployed in many--if not most--Fortune 1000 enterprises, and wide-scale roll out of storage virtualization Treating storage as a single logical entity without regard to the hierarchy of physical media that may be involved or that may change. It enables the applications to read from and write to a single pool of storage rather then individual disks, tapes and optical devices. is gathering steam. Every major storage vendor in the market has blessed the concept of virtualization, despite some initial foot-dragging on the part of some hardware manufacturers. While it may have seemed a threat initially to homogeneous The same. Contrast with heterogeneous. homogeneous - (Or "homogenous") Of uniform nature, similar in kind. 1. In the context of distributed systems, middleware makes heterogeneous systems appear as a homogeneous entity. For example see: interoperable network. storage networks, most vendors now realize virtualization will instead provide a whole new market for flexible, powerful storage applications in businesses large and small. Virtualization has to come of age. Trends to Date Until recently, the storage community has debated the pros and cons pros and cons Noun, pl the advantages and disadvantages of a situation [Latin pro for + con(tra) against] of two primary storage virtualization architectures: in-band and out-of-band. Without going into too much detail, a brief overview of each--in terms of infrastructure and scalability, with strengths and weaknesses--should be mentioned. In-band solutions rely heavily on hardware (virtualization servers) running as intermediaries between the host network and the storage pool. All transactions funnel through this appliance, creating limitations on potential performance and introducing complications in terms of availability. The in-band appliances must have a "fail-over" capability, otherwise, funneling all I/O (Input/Output) The transfer of data between the CPU and a peripheral device. Every transfer is an output from one device and an input to another. See PC input/output. I/O - Input/Output through a single appliance creates a single-point-of-failure. Additionally, management of these appliances is not always capable of being centralized cen·tral·ize v. cen·tral·ized, cen·tral·iz·ing, cen·tral·iz·es v.tr. 1. To draw into or toward a center; consolidate. 2. , so multiple points of management must be accommodated. In contrast, out-of-band solutions, while they may deploy some distributed hardware components to handle metadata (1) (meta-data) Data that describes other data. The term may refer to detailed compilations such as data dictionaries and repositories that provide a substantial amount of information about each data element. , are primarily software-based. The data flows directly from the server to the storage subsystem The part of a computer system that provides the storage. It includes the controller and disk drives. See storage system. and, therefore, performance and availability are not impacted. Because of the out-of-band or distributed architecture, centralized management is typically offered. A Need to Refocus Verb 1. refocus - focus once again; The physicist refocused the light beam" focus - cause to converge on or toward a central point; "Focus the light on this image" 2. on Solutions Focusing on the pros and cons of competing virtualization technologies See VT. See also virtualization. has been, perhaps, a necessary evil in a market where 24X7 data availability Refers to the degree to which data can be instantly accessed. The term is mostly associated with service levels that are set up either by the internal IT organization or that may be guaranteed by a third party datacenter or storage provider. and data protection rightly rule supreme. Open discussion of the "best" possible means to provide heterogeneous storage networks is a good thing for the industry and a great service to customers sitting on the sidelines On the sidelines An investor who decides not to invest due to market uncertainty. on the sidelines Of or relating to investors who, having assessed the market, have decided to avoid committing their funds. . Switch-based virtualization will, no doubt, raise the level of debate significantly yet again. But both in-band and out-of-band solutions have been "field-tested" extensively over the last two to three years; furthermore, one technology may not be appropriate for all customer solutions. Intelligent switches will no doubt play an even greater role in the coming months. Meanwhile, many shops are living and breathing virtualization on a daily basis in full production environments and garnering huge savings on storage management and deploying superior storage protection. However, many other businesses have a pent-up demand for these same solutions, for applications that only a virtualization platform can provide. As a result, the storage industry needs to refocus now on the tremendous customer benefits virtualization makes possible. Customers are seeking working, proven solutions to many of the following problems. * Better utilization of disk resources * High availability Also called "RAS" (reliability, availability, serviceability) or "fault resilient," it refers to a multiprocessing system that can quickly recover from a failure. There may be a minute or two of downtime while one system switches over to another, but processing will continue. and disaster recovery * Snapshot (1) A saved copy of memory including the contents of all memory bytes, hardware registers and status indicators. It is periodically taken in order to restore the system in the event of failure. (2) A saved copy of a file before it is updated. and replication In database management, the ability to keep distributed databases synchronized by routinely copying the entire database or subsets of the database to other servers in the network. There are various replication methods. for rapid recovery * Central management of storage Point solutions exist for many of the above storage demands, but customers are increasingly under pressure to consolidate management and applications to save time and money. This consolidation--not just of storage itself, but also of storage management and applications--is what is really driving the new storage market. The only solution that addresses all of the IT concerns listed above--supporting centralized management--is a solid virtualization product providing the bedrock for heterogeneous storage applications. Only virtualization can solve mirroring enterprise-wide data from disparate data sources, snapshot and replication of entire storage pools or the automatic provisioning See automated provisioning. of disk across multi-vendor storage arrays. And there appear to be no other answers on the technological horizon. But buyers beware--not all virtualization is created equally. Corporations seeking to finally consolidate storage and storage management must consider the underlying implementation of a virtualization product, and how disruptive--or non-disruptive--it is to their environment. They must ask the hard questions. Is the technology mature? The length of time a product has been in real-world production environments is a key indicator of product maturity, as well as the variety and size of the implementations at real customer sites. Since one-size-never-fits-all in enterprise storage, a virtualization solution should be flexible enough to handle SQL SQL in full Structured Query Language. Computer programming language used for retrieving records or parts of records in databases and performing various calculations before displaying the results. , Oracle, Exchange, SAP and a multitude of other data sets and network configurations. Does it deliver real value? Virtualization alone gives you nothing more than volume management and disk sharing. Does the platform provide a solution for snapshot, replication and mirroring? Does it lend itself to central management and control? Is it scalable over the long term? What virtualization can do for one application on one network segment may be significant, but can the enterprise leverage the technology across the entire network and all storage? What is the true cost of going from a pilot program to a full-blown enterprise-wide solution? Does it accommodate future technologies and applications easily? What impact will it have on network performance? Will it accommodate growth in storage and provide the best possible platform for continued disaster recovery and business continuity for the majority of the enterprise applications? Business continuity is spreading its wings. The trend toward Information Lifecycle Management Information Lifecycle Management refers to a wide-ranging set of strategies for administering storage systems on computing devices. Specifically, four categories of storage strategies may be considered under the auspices of ILM. (ILM) has helped companies realize that it is not just their core business data that should be protected by disaster recovery implementations. E-mail, document processing Processing text documents, which includes indexing methods for text retrieval based on content. See document imaging. , customer response management and other mid-level applications are increasingly recognized as mission-critical within the organization and, therefore, are candidates for DR and Business Continuance The adjournment or postponement of an action pending in a court to a later date of the same or another session of the court, granted by a court in response to a motion made by a party to a lawsuit. processes. Intelligent Switches: the Next Virtualization Platform As storage switches gain intelligence and additional processing capabilities, the natural place for virtualization to reside may be in the fabric of the storage network itself. This approach is, in reality, both in- and out-of-band in nature, and may be more attractive to customers who will be shortly upgrading their storage switches to higher bandwidth and the latest technology. Besides providing an efficient and economically upgradeable location for hosting virtualization engines In computing, a virtualization engine is the concept of giving a holistic view of all the resources in the entire network infrastructure. The holistic view is independent of the physical data storage devices and their geographic location. , intelligent switches may be the best place to run application solutions such as snapshot and replication across an entire enterprise with no performance impact on production servers and applications. Switch-based storage applications will lend themselves more easily to centralized management as well. Upgrades and new storage applications are applied at the switch level, making introduction to the storage network less disruptive than other approaches. In addition, software residing on a switch is a magnitude less expensive than either deploying multiple in-band servers or point solutions on the host network. Multiprotocol Storage Another force is coming quickly to bear on storage networks over the next two years: multiprotocol storage. At first examination, the introduction of iSCSI and FCIP (Fibre Channel over IP) A protocol for tunneling Fibre Channel data across an IP network. Fibre Channel was designed for local storage area networks (SANs), but FCIP extends the distance to remote locations via any IP network. See Fibre Channel, iFCP and IP storage. into the traditional Fibre Channel SAN may seem to argue against the use of virtualization and even pose a threat to Fibre Channel (FC) in general. In reality, the adoption of multiprotocol storage may actually enhance the value of Fibre Channel and virtualization. In the case of iSCSI, many enterprises may now reach out and connect clients and servers to the FC SAN that were not previously attached due to cost concerns. This enables greater utilization of the storage assets on the SAN and provides higher quality of service to these "stranded" servers, iSCSI enables businesses to attach segments of their data networks to FC devices as never before, making not only the devices more valuable and mission critical, but also enhancing the role of Fibre Channel as a central management point and natural location for storage applications. FCIP may now make interconnection in·ter·con·nect v. in·ter·con·nect·ed, in·ter·con·nect·ing, in·ter·con·nects v.intr. To be connected with each other: The two buildings interconnect. v.tr. of SAN islands and remote replication of data affordable to more and more businesses. The cost of remote data transmission alone has hampered wide-scale implementation of solutions such as remote backup, remote snapshot and remote mirroring for most customers. With even more data, from ever-more disparate sources to contend with, switch-based or out-of-band-based virtualization can play a key role in protecting and enhancing the management of this new source of data. A Focus on "Solutions" While there will always be a debate over which implementation or architecture is the most robust, there is growing evidence that the market is moving beyond a technical discussion and is now focusing on a discussion of solutions. This is encouraging, as more users are deploying virtualization architectures to enable heterogeneous snapshot, replication, migration and mirroring solutions. "Predicted" benefits are now being demonstrated and deployments are expanding. The adoption of intelligent networks--pushing more intelligence into the switches--will continue to escalate es·ca·late v. es·ca·lat·ed, es·ca·lat·ing, es·ca·lates v.tr. To increase, enlarge, or intensify: escalated the hostilities in the Persian Gulf. v.intr. the adoption of virtualization and data protection applications. These intelligent switches will make many of these applications ubiquitous and available to all servers and storage devices within the storage network. To take full advantage of networked storage, users should make sure the architecture will allow for scalability on three levels. These three parameters ensure complete scalability. * Addition and deletion deletion /de·le·tion/ (de-le´shun) in genetics, loss of genetic material from a chromosome. de·le·tion n. Loss, as from mutation, of one or more nucleotides from a chromosome. of servers and storage devices as needed as needed prn. See prn order. * Addition and deletion of network protocols and technologies as needed * Distribution over geographies as needed The future of virtualization and intelligent data protection is very promising, if the industry continues to focus on customer solutions that demonstrate tangible value.
Virtualization Scalability Primary Infrastructure
Method Applications
In-Band Based on Storage pool Primarily
number of central management hardware
virtualization mirroring, snapshot
appliances and replication
Out-of-Band Up to network Storage pool Primarily
capacity central management software
mirroring, snapshot
and replication
Virtualization Performance
Method
In-Band Dependent on
in-band hardware
appliance and fabric
Out-of-Band Dependent on fabric
Kevin Liebl is vice president of marketing at StoreAge (Irvine. CA) www.store-age.com |
|
||||||||||||||||||||
Printer friendly
Cite/link
Email
Feedback
Reader Opinion