The post How does point in time recovery work? appeared first on Over the Wire.

Although they may be the last place you want them to occur, backup systems are not immune from problems. To protect against failure, backup providers will often offer options for data replication, wherein archived data is duplicated across two or more disks. In highly sophisticated systems this can be done across multiple data centers to afford businesses site redundancy.

However, these options will cost more due to the extra disk space required so businesses will need to balance up the increase in expenditure with the chance of failure, and the bottom line for the consequences of failure, which can be substantial should a disk be unavailable when a restore is required.

2. Global Deduplication

Deduplication refers to the process that backup software goes through to prevent the archiving of multiple copies of the same pieces of data.

Global duplication simply means that this process happens universally across all of the different data sources (eg. separate servers, sites, etc.) rather than discretely within each individual source. Because most businesses have multiple copies of the same files spread across multiple sources, global deduplication is often capable of saving an enormous amount of disk space. As such it has become relatively standard feature in backup and recovery services.

The differentiating feature that businesses will want to keep an eye out for nowadays is whether or not the deduplication is a source side or destination side process. Destination side deduplication is the standard, so if the provider hasn’t specified anything they are probably using this. Source side is much less common, but comes with substantial advantages over destination side processing.

In a source side model the deduplication occurs before the information leaves the client’s environment, meaning that the business will have a substantially reduced load on their network bandwidth. In many cases this will mean that businesses can make backups more frequently, and in spite of peak load periods.

3. Ease of Administration

Even highly sophisticated backup and recovery systems need to be maintained. Providers typically offer two solutions for this: self managed maintenance plans, and fully managed maintenance plans.

As the name would suggest “fully managed” places the onus on the provider to ensure the smooth operation of the backup and recovery process, so clients should make sure ahead of time that the agreement stipulates an appropriate level of support.

It must however be kept in mind that even the best fully managed solution is only as good as the infrastructure on which it is deployed. In contrast, a self-managed service places the onus on the business. There is nothing inherently problematic with this, but businesses will need to make sure that their IT staff have the time to maintain the backup systems appropriately, and factor these time costs into their budgets.

4. Business Continuity Planning

Business Continuity Planning (BCP) is a consultation service offered by the provider in which they will design a backup solution to suit the requirements of your business. By utilising BCP your provider will help you to determine what technologies are and are not necessary for your organisation to use.

This will be based on your desired Recovery Point Objectives (RPOs), which stipulate the recency and frequency of the data updates (ie. every 24 hours for the last month, every 8 hours for the last week, etc.) and your desired Recovery Time Objectives (RTOs), which stipulate the maximum time requirements for a system restore.

Although for some organisations BCP is not a top priority, it has the potential to change the costs of your backup solution drastically as the service is entirely reliant on the expertise and experience of the provider.

5. Fast Recovery Times

Whilst having a detailed backup of your company’s data is beneficial, it’s the time in which you can restore this data (in part or in full) that can have a significant impact on your businesses ability to recover from a loss.

The traditional tape based onsite backup methods often do not facilitate the ability to quickly restore the full system, and if the hardware is no longer available this process can take days if not weeks to completely recover.

One of the best recovery speed indicators are Recovery Time Objectives (RTOs), which stipulate the maximum time a restore is likely to take. Because recovery times are dependent on the infrastructure and levels of support in place, as well as the volume of information to be restored, RTOs will vary for each individual business and are therefore not stated until a consultation has taken place.

Several rule of thumb indicators for fast recovery times that businesses can look for are listed below:

• The availability of data replication which can improve recovery times by minimising complications.
• What the SLA’s of the provider stipulate with regards to service availability and response time.
• The available bandwidth on the connection between the business and the backup archive for expediency of transferring backups.
• The level of availability and reliability of the data center(s) in which the backups are stored. Data centres with a Tier III or higher rating are preferable.
• The ability to house the backup data in close proximity to the restore location. Combinations of BaaS/RaaS and Infrastructure as a Service (IaaS) are ideal, however where the hosting facility is unreliable this increases the risk posed by a site outage.
• Providers which offer Business Continuity Planning will provide advice on the best way to meet a business’s required RTOs.

The post Five Things Every Business Should Look For in a Backup Solution appeared first on Over the Wire.

The first important item to take note of when considering backup systems is that in recent years traditional tape based backup systems have become obsolete in the face of cloud based models, which eliminate the labour intensive process of changing and storing tapes. In addition, cloud based models also eliminate the high capital expenditure required for the tapes themselves, for which the number of days of backup is determined by the number of tapes your organisation is willing to purchase.

Tapes also have the drawback of physical vulnerability; even when stored in remote locations the tapes are vulnerable to hazards both in transit and by virtue of being a single copy. Moreover, because tape models use a sequential backup system if one of the tapes in a sequence is corrupt or damaged you may be unable to recover your data.

With most traditional onsite methods still including the use of data tapes, the process of ensuring that reliable backups are maintained, as well as decreasing the time requirements for getting the data back can be quite challenging. However, businesses can now look to leverage the high performance capabilities of WAN technologies to extend beyond the onsite model by using a cloud based solution.

Cloud BaaS systems are more reliable as they store the information in data centres, offering significantly higher levels of virtual and physical security than would otherwise be available. Additionally, the infrastructure holding the data doesn’t require physically moving locations, eliminating transport as a potential point of failure.

The second important item to note when considering backup systems is the difference between Backup as a Service (BaaS) and Recovery as a Service (RaaS) models. Typically BaaS models only allow for the backup and storage of the actual data. This means that in the event of wide scale data loss or system failure you will be able to recover the data, but nothing more. So in order to get your systems operational again you will have to undergo the lengthy process of restoring your operating environment before you will be able to begin recovering your data. Depending on the complexity and size of the affected systems this process can take anywhere from a few hours to several days to complete, with the time for a full environment rebuild often not considered in the recovery plan.

Recovery as a Service (RaaS) models differ from BaaS models in that the entire operating environment is backed up, not just the data. As these backups can then be restored on virtual infrastructure (IaaS) the whole environment can be recovered and spun up onto the new hardware in a fraction of the time it would take to rebuild the servers in a BaaS setup.

More sophisticated Recovery as a Service platforms can be offered as combinations of RaaS, Disaster Recovery as a Service (DRaaS), and Business Continuity Planning (BCP) rolled into one. Businesses should always be sure to check with providers whether this level of service is available.

The post Not All Backups Are Created Equal appeared first on Over the Wire.