Cloud-based computing is an information technology model by which resources, software, and data are provided as an on-demand service. The best metaphor is that cloud-based computing seeks to make information technology available in the same manner that your utility company makes electricity available—with a simple wire without your having to understand the details of how it’s provided.
For Backup and Recovery, Clouds are often used for Disaster Recovery as a Service (DRaaS), and include use for the storage of the data (Cloud Storage), as a location in which to recover servers and data (Cloud Backup), and to scale (or add resources to existing servers or applications) to aid in avoidance of an outage.
Cloud storage simply refers to the use of the cloud as a replacement for some NAS (Network Attached Storage) centralized storage—some vendors advertise their cloud storage as “your file server in the cloud.” Cloud backup refers to the use of the cloud as a mechanism to protect data.
Pure online backup is offered by an ever-growing number of vendors. And most of these online backup vendors do a credible job of backing up a few tens of gigabytes of data. The real problem with online backup is recovery. You might not care that it takes a month or more to ship your first terabyte up through the Internet to the online backup vendor. However, most people don’t have a month to wait for that terabyte to be downloaded back from their online backup vendor when a hard drive or a complete system is lost.
The answer to rapid recovery while taking advantage of cloud-based computing is an on-premise backup appliance that can quickly serve to backup your information technology infrastructure and serve as a gateway point to a cloud-based disaster recovery service.
An on-premise appliance offers not only local backup but also offers dedicated in-flight deduplication. Once the cloud is loaded with your initial set of data this in-flight deduplication not only compresses your data but also deduplicates that data before it is sent upstream to the cloud. This is done with no impact to the clients (i.e., servers, PCs, workstations, and notebooks) that are being protected by the on-premise backup appliance. It means that once your initial set of data has been uploaded, much more data can be protected than can physically move over the WAN in any given time. Thus it is possible for hundreds of gigabytes and more to be protected and kept synchronized with the cloud each day with a relatively small amount of bandwidth.
Advantages of Cloud Solutions
The fundamental advantages of cloud-based computing are
- Reducing costs. This advantage is primarily accomplished by eliminating capital expenditure and reducing operational expenditure. For multi-tenant public clouds, cost reduction is also possible by the pooling of information technology resources in order to be able to meet peak demand with effectively less information technology resources.
- Ability to quickly scale. This advantage refers to the fact that because cloud-based computing typically has excess capacity, that this excess capacity may be provisioned quickly and in a highly responsive and nimble fashion to those users requesting it. To put this more simply, there is no wait for new hardware and/or software to be ordered and installed.
- Well-defined service level agreements resulting from measured service. Typically cloud-based computing services are priced to some degree based on usage characteristics; thus the service level agreements associated with everything from availability to performance are more rigorously captured than alternative solutions.
Drawbacks of Cloud Solutions
Of course there are drawbacks with respect to cloud-based computing as well; these include
- Performance and latency. WAN performance has not and will not keep pace with microprocessor, memory, disk, and LAN performance. Any cloud solution must take into account not only the gap between on- and off-premise performance, but must also address the fact that this gap will only widen over time. This issue affects both private and public clouds equally.
- Security and privacy. Security and privacy, for both at-rest and in-flight data, must be taken into account. Different industries have different requirements so it’s important that security and privacy be addresses in a flexible manner. This issue affects primarily public clouds.
- Reliability and availability. The information infrastructure that constitutes the cloud as well as the connection to the cloud must be both reliable and highly available. This issue affects both private and public clouds equally.
- Manageability. Because clouds can be somewhat amorphous in nature to the end user and because users perceive a loss of control when using clouds, the manageability of the cloud is very important. In private clouds, this is typically less of an issue since the entire infrastructure can be managed by the one tenant of the cloud. In public clouds, manageability has to be a key component of the architecture since otherwise a user can end up with a lowest common denominator solution. Monitoring. Trust is the single biggest issue raised by consumers of cloud-based services. Monitoring is incredibly important as one way of helping establish trust. This is a major issue for public clouds—and has also been called out as an important issue in private clouds as well.
Does Your Cloud Have a Silver Lining?
The level of hype surrounding cloud-based computing has never been higher. The purpose of this paper is to help you understand the advantages and the drawbacks of cloud-based computing—particularly as it applies to business continuity and disaster recovery.
Thank you for your interest in Does Your Cloud Have a Silver Lining?