News
FalconStor, Beyond virtualisation: putting a good idea to good use
November 2007 by FalconStor
Over the past few years virtualisation has seemed to suffer from the after effects of ‘too much buzz’: everybody talked about it, but few said the same thing. Disk vendors claimed virtualisation lived on storage arrays, while other vendors took existing, host-level, volume management packages and gave them a “virtualisation” label. And some vendors championed virtualisation in the storage network, the model that is the most complete: by interposing a layer of intelligence between servers and storage, network-based virtualisation offers the most freedom from operating system constraints and vendor-specific storage devices.
The confusion that has resulted is unfortunate because virtualisation truly offers significant benefits to storage data management. It offers a layer of abstraction above the physical layer of disks and spindles; it also unifies everything and makes it all look the same - in short, it provides a utility model for storage. Storage virtualisation also provides a way to bring storage services into the network, without any reliance on operating system or storage-specific tools. This is the true value, and is what many end users have failed to realise among all the competing claims.
Today, virtualisation is heading decisively towards services. In fact, services are what virtualisation was always about, and is why the network-based virtualisation model, with its greater flexibility, offers the most value. The leading virtualisation solutions comprise an ever-growing suite of critical storage tools to enhance data performance, availability, and recoverability, including:
• Data mirroring: duplicating copies of data across different storage devices
• High Availability: with virtualisation, the server is automatically redirected to a failover system, without any outage and without any manual intervention.
• Data replication: long-distance data movement over IP, periodic or continuous, for disaster recovery
• Tape backup enhancements: bridging the gap between disk and tape, allowing high-speed, minimal impact backups to take place.
• Point-in-time copies: Enterprises always need copies of data, for testing, development, reporting, for example
• Point-in-time data recovery: moving near-term data (data up to several weeks old) to disk resources, rather than tape, for super-quick, reliable restore
• Application integration: by integrating storage services with the applications they serve, far greater value can be derived. Operational headaches also reduce.
There is a second new direction as well: providing the storage services outlined above without having to virtualise data at all – i.e. without having to migrate data into a new, proprietary format. Put another way: providing the key benefits of virtualisation without the perceived risk factors.
FalconStor offers a means to accomplish this using its patent-pending Storage Service Enabler technology, a part of its IPStor software platform. The Service Enabler allows end users to "import" existing disks and LUNs into IPStor without migrating data, yet still taking full advantage of storage services such as mirrors, replication, point-in-time recovery, ease of assignment, etc. This provides an extremely fast means to add storage services to an existing, multi-vendor storage infrastructure. It also eliminates all fears of being locked into a proprietary virtualisation paradigm for a simple reason: there is no virtualisation. Any disks imported by IPStor can simply be removed from the system and reconnected directly back to the hosts they came from, without any impact on the data structure. However, as long as the disks are managed through IPStor, the end user gains the benefits of a unified view of disparate resources.
With products such as this, the industry is truly moving beyond virtualisation to open, standards-based storage services platforms that still provide the features which have, until now, been the preserve of proprietary schemes. The history of technology of all kinds proves one thing: ‘open’ is better.
The next step – putting virtualisation to practical use
Ultimately, virtualisation is just a technique used in the newest state-of-the-art data protection solutions offered by Falconstor. Today, Falconstor is ready to take virtualisation technology to the next level, with its Single Instance Repository (SIR) and Continuous Data Protection (CDP) solutions representing a revolution in data protection.
FalconStor SIR uses redundant data elimination (RDE) technology to simplify long-term data protection, intelligently eradicating data redundancy by focusing on the characteristics of the data itself. This dramatically reduces the storage media requirements and keeps costs for long term data retention and archiving to the absolute minimum. Additionally, FalconStor SIR vastly reduces the costs and risks associated with replicating data to recovery sites or offsite archives. The benefits of FalconStor SIR can be organised simply into the following “Six Ss” – Savings, Scalability, Speed, Security, Synergy, and Simplicity.
CDP is the process whereby data is captured and replicated to a separate storage location to ensure that a set of critical data is always available. FalconStor CDP solutions enable organisations to meet their stringent Recovery Point Objectives (RPO) and Recovery Time Objectives (RTO) by facilitating simple deployment of nearline storage within a data centre or disaster recovery (DR) storage at a secondary site to protect production servers against human error, application failure, disk failure, and site outages while minimizing data loss and downtime. With CDP protection, in the event of any outage, planned or unplanned, administrators can simply roll back data to any known good point-in-time.
Virtualisation is a key component for today’s changing storage environments, enabling us to change the way we approach data protection. Only a few years ago, the main focus was on backup itself. Emerging technologies were all about the assurance that a backup could be completed within the available backup window. This concept has evolved into a restore discussion: until recently, everyone agreed that it is not the backup that is important, it is the restore that has to be available.
But today, we realise that even the restore is not enough. A restore just means that data can be copied back from tape onto disk. Today we need more then that, we need the application to be back online within the shortest period of time possible. Recovery is what really counts.
It is virtualisation that allows us to get more out of the limited resources that are available, while meeting new requirements for system and application recovery. Virtualisation is not a goal as such, but it is certainly the key technology that is enabling the revolution in data protection.
Related articles:
