There are essentially two forms of backup types, one for disaster recovery for when your business has lost its computers through fire or theft, and the other, is long term archival. This latter form of data backup is usually to either meet legislative or governing body requirements, or to retain essential client or project data for future reference.
In any data backup system, at least two copies of the data should be retained, preferrably more. The reason for this is, should a failure occur during an update, then the most recent backup may then be unusable due to corrupt, missing or partial data. By at least cycling two or more copies of the data, there is a lesser chance of this happening.
Data Backup For Disaster Recovery
This is where you backup not only your data, but also the settings of your computer system. Typically for most small business this is the route taken for backing up their server, as should you loose the settings, then all the users passwords and permissions will be lost, and even though you may have the data, it may not be readily available without the passwords or permissions.
In taking the full backup of the computer (or server), you can use Microsoft Windows backup (previously Windows NT Backup), but I would strongly recommend against doing this unless you are already running this server in a virtual machine environment, as, in times of urgency when trying to source a replacement physical server after a building fire or theft, it is very difficult to obtain like for like hardware, and the backup simply will not restore to the new hardware due to differences in chipsets, drivers, disc volumes etc. The way around this is to use backup software with built in capabilities for restoring to an alternative hardware type. This type of software is called “bare metal recovery” with “flexible hardware restore”. If you have already virtualised the server machine, then this issue has been nicely side stepped.
We would strongly recommend businesses invest in this type of data backup and recovery software, especially for aging servers. The following brands of software all have this capability: Acronis, Ghost or Disk Image. We have had considerable success with Acronis and we recommend this for business backup disaster recovery use.
Data Backup For Archiving
Now, data backup for archiving is a different matter altogether from above, as it may be necessary to keep data, but not server settings as above, for an extremely long time. Typically, a single copy is taken, then stored off site (off line) perhaps in the cloud, however, there is nothing stopping multiple copies to be taken, perhaps one local for quick restores, and one remote in the cloud somewhere. The problem with this type of backup is that you can easily get a data explosion, ie, if you are storing 200Mb of data for a typical small business, and need to keep a monthly backup, then with 12 straight copies, annually that is around 2.5Tb of data. Keeping all these copies for six years would be 15Tb of data! A huge amount, and a large cost burden to any business. In analysing what is stored in the archive, clearly most of the data from one month to the next may not be changing, and a complete copy over and over again as just described would be wasteful of cloud disc space given a file would be stored 72 times in that 15Tb of data.
To overcome this data storage explotion, is it necessary to use software that will simply add the changes to the initial backup seed. The software initially will write a full copy of your businesses essential data to a file store, then daily add to it. This file store should then be replicated elsewhere or rotated on a regular basis. For software that is good at doing this, try the inbuilt Windows Backup, or for a commercial version Acronis or for free FBackup.
Windows servers can take regular snapshots of data folders, and it is possible to roll back to a previous version when necessary, but, there is a limit to the volume that can be effectively stored on a Windows server. This shouldn’t be relied up as an archiving system.
Cloud Based Data Storage
There are other permitations of the above data archiving theme offered by companies in the cloud, and some of them also incorporate data sharing between different users and different devices. I do however urge some caution with these services, as there may be no local copy for a quick restore, and there needs to be careful management to ensure that you do not incur significant costs. Not only that, you cannot guarentee where your data is housed, and this may be an issue for some European based companies due to Safe Harbor. These cloud based services are only suitable for backing up files, and not for replicating disc images to provide a bare metal recovery after a complete disaster. Examples of this type of service are: Dropbox, Google Drive, One Drive (previously Sky Drive), and also iCloud.
The Best Data Backup Options
Well, after discussing the above, a small business with their own server actually needs each of the above, for different reasons. They will require a regular snapshot of the server in case of a disaster, and, additionally will require archiving versioning for when a user needs to return to a previous version of a file, and perhaps a DropBox type of service to share files with other clients.
In implementing this, it is clear that two stores are required, one local to update or restore quickly, and one remote. The remote version is updated by the local version. This is where RKive.IT comes in. We offer the storage platform which we call RKive.IT that is independent of the software used to perform the backup, and the RKive.IT system will automatically synchronise the two copies, local and remote.
For more information, jump to the following pages within this site: