Data restore is the process of copying backup data from secondary storage and restoring it to its original location or a new location. A restore is performed to return data that has been lost, stolen or damaged to its original condition or to move data to a new location.
There are several circumstances that lead to the need for a data restore. One is human error, where data is accidentally deleted or damaged. Other circumstances include malicious attacks where data is exposed, stolen or infected; power outages; manmade or natural disasters; equipment theft, malfunctions or failures; or firmware corruption.
Data restore makes a usable copy of the data available to replace lost or damaged data and ensures the data backup is consistent with the state of the data at a specific point in time before the damage occurred.
Preparing for a data restore
Data restore is part of the overall data management process and is contingent on having a system in place to produce a good copy of the data being protected by traditional backup, snapshots or continuous data protection (CDP). Without a reliable protection copy, there is nothing usable to restore.
To ensure a reliable data backup version is available to restore, it's necessary to test the restore process and the data recovery tools used. Protection copies should be randomly checked at various points in time to ensure they meet recovery point objectives (RPOs). Data being restored must be readable, consistent with a chosen point in time and include the information needed to comply with RPOs, recovery time objectives (RTOs) and other service requirements.
In addition, all applications must be checked before doing an actual data restore to ensure they will be able to use the restored data. That means the software used to format the data must be available, and security certificates, permissions, access control and decryption must be applied correctly.
Methods of restoring data
Where backup data is stored will affect the ease with which it can be restored.
Hard disk drive (HDD) backups provide a quick data restore because it's easy to locate data on disks, and the systems often live on site. For this same reason, HDDs are more secure than off-site tape and cloud backup. However, disk systems cost more than other data backup and restore methods; costs include the power needed to run both the disk systems and the cooling systems they require. HDD backups are best for data that changes frequently and requires a short recovery time.
Tape backup systems provide high-capacity storage at a lower cost than that of HDDs. But even with the latest technology, tape still has a longer recovery time than disks or the cloud, and that timeframe expands when data is stored off site. Tape libraries require ongoing management and testing to ensure data is accessible when needed.
In this video, George Crump, president and founder of Storage Switzerland, discusses data restore issues that arise with cloud service providers.
Cloud backup requires enterprises to send a copy of data over the corporate network or the internet to an off-site server owned by the enterprise or hosted by a service provider. When it's time to restore that data, it must traverse the same path, which can take time due to network bandwidth limitations. For this reason, cloud backup and restore is generally used for noncritical data with a longer RTO.
With cloud backup, it's easy to add capacity as data backup needs grow. In addition, costs are lower, particularly when using a cloud provider, because organizations don't have to buy and maintain backup software and hardware. Using a third-party provider also reduces the workload on the IT department. However, as data volumes grow, cloud backup costs rise.
Data restoration techniques
The approach used to restore data depends on what information was lost or damaged, how much data was affected, how the incident happened, the software used to create the data backup, the backup target media and other factors. Some backup software enables users to restore lost files themselves. Data recovery software and services can retrieve accidentally deleted files that aren't backed up from the hard drive. More complicated data loss or damage requires IT to restore backup files from disk, tape or other backup media using various techniques, including:
- Instant recovery, also known as recovery in place, redirects a user's workload to a backup server, eliminating the recovery window. Users get almost immediate access to a snapshot restore point of their workload where they can work while IT manages the full recovery and data restore in the background. Once that process is complete, the user's workload is redirected back to the original virtual machine.
- Replication provides even faster, near-instant access to data; however, data backup with integrated replication often lacks a product that provides historical recovery and isn't a true backup capability.
- CDP is when data is backed up using snapshots taken every time the data changes. This approach lets data be rolled back to any point in time; however, it comes at a price because it can tax a system's CPU and require a lot of storage to accommodate heavily updated data.
- Near-CDP is when snapshots of changed data are taken at set intervals and changes are then consolidated over a longer period of time. This approach cuts down on the total amount of storage required to accommodate backed-up data compared with full-fledged CDP.
- Traditional backup is when data is stored on HDDs or magnetic tape either locally or at a remote location. Traditional backup is most useful when a major hardware or site disaster occurs. It lacks the scalability and efficiency of other methods, but it is a better long-term approach for data retention and restoration.
Mobile backup and restore
Backing up and restoring mobile data from smartphones, tablets and laptops poses specific challenges. Traditional backup software often assumes that devices being backed up have a permanent location, consistently good connections to the corporate network and adequate bandwidth. But mobile devices frequently don't have access to these capabilities.
Enterprise file sync-and-share (EFSS) services protect data on mobile devices by copying files to the cloud or to on-premises storage. EFSS enables users to access these files on other desktop and mobile devices, but it is not true backup and doesn't allow for the rollback of data to a particular point in time should the device fail, become lost or stolen, or if data on it is damaged or destroyed.
Most Android and all iOS devices have native, image-based backup, but that leaves the responsibility for backing up these devices with users. An endpoint backup product that supports mobile devices and incorporates file sync and share is one way to handle this. As with all enterprise data backup and data restore procedures, the key to smooth data restoration on mobile devices is to have a consistent, tested backup process and data recovery tools in place so data can be restored quickly and easily when necessary.