The way we look at backups – the simplest method



Flat File / File Level Backup Versus Full System State/Image Backup

There are a lot of methods in this current day and age to backup you Critical data, but what must be taken into consideration is the percentage of your business that is dependent on your IT infrastructure and if your backup method meets your Business requirements. When it comes to backups in the past, due to the cost of backup software/hardware and target Destinations such as Tape & Disk, you were limited to only protecting your critical servers. On that note the backup method chosen was often the cheapest round trip solution. Even still, companies only backup critical servers, some only the critical data on those servers through file level or flat file. The changes in technology and the growth in cloud technology has led to many organizations planning DR strategies as they are now affordable. This introduces the RTO & RPO as key features in any DR strategy.

What Defines Critical servers? Here is a simple explanation of ‘mission critical’.

Mission critical refers to any factor of a system (equipment, process, procedure, software, etc.) whose failure will result in the failure of business operations.

Before I go into comparing the 2 backup methods, what must first be understood is that backups are important, so make sure you choose a solution that will work for the needs of your environment. But more important in my opinion is the restore capability. Restoring your data is even more crucial than having it backed up. You need to ensure that the solution that you use can easily restore critical data when needed in your environment.

What you need to define is whether or not you need to be able to restore a critical business service or just critical data when lost. If you require restoring a service, then you should also consider planning your backup strategy around your current or future disaster recovery strategy.

Good backup infrastructure is the foundation for any disaster recovery plan.

Choose a backup strategy that supports your disaster recovery requirements for recovering from a disaster.

Flat File & File Level Backup Methods

File systems store files and other objects only as a stream of bytes, and have little or no information about the data stored in the files. Such file systems also provide only a single way of organizing the files, namely via directories and file names.

A flat file is a file containing records that have no structured interrelationship. The term is frequently used to describe a text document from which all word processing or other structure characters or mark-up have been removed. Typically an image file like .MDF .CTK .BIN etc…

When referring to Flat File & File Level backups, this generally entails a scheduled job that backs up specific individual files (either Flat in the case of a Database servers DB files or file and folder level on files server or Application server with a critical file store).

The main benefits of this is to backup a small amount of data on a large volume rather than the entire volume. For example, consider a backup consisting of 10 gigabytes of critical database files that reside on a 120 gigabyte hard disk. The other 100-odd gigabyte could contain OS and other non-crucial data. Another advantage is that file-level backups allow for individual file restores, which are useful when you have just a few files to restore and do not need to restore all the data just to restore the individual files.

However, consider the IT Engineers installing service packs to production type software such as Exchange, SQL, or other software that is critical to your business, or the installation of new hardware/software and you have a full system state failure to your business critical servers. Your course of action with only having file level/flat file backups in place will be (this will calculate to be your RTO):

  1. OS rebuild
  2. Application install
  3. Patching (in some cases you can’t restore system databases unless the patch level is of the same build to when they were backed up)
  4. Then you can begin to restore individual flat files to your new server

So file-level backups have their limitations. In other cases, operating system files or locked files will not be backed up, and data discovered in the volumes is not structured. That means backup will not be application consistent. Therefore, this backup type is not suitable either for restore in the event of a disaster or for granular restores within applications such as databases, email, etc.

Building out a disaster recovery plan on this method is not fail proof and it’s not flexible solution. It’s very manual and will involve a lot of maintenance & fail over testing. Unless you incorporate Arcserve Replication and High Availability into your file based strategy you don’t have continuity (I will post more on RHA in upcoming posts).

Full System State/Image Backup

Image/System State Backups in the form of an agent or agentless on virtual or physical allow you to select an entire drive, partition, or entire machine which backs up the entire selection that you have selected. This backup then covers your system data crucial platform for your critical company service or DATA. Examples of these situations are when your hard drive dies, your Windows Guest will not boot up or is corrupt or your servers are stolen. “This is Africa “. Or as mentioned earlier, technical team update rollout goes horribly wrong, or another major disaster that requires a complete rollback of the entire machine. The advantage to image-based backups is that all of the information can be collected in a single pass, providing an updated bare metal restore (BMR) capability with each file-based backup.

You can restore back to dissimilar hardware. It allows for file-level restores from the backup image. You are able to recover servers remotely across wide area networks (WANs) or local area networks (LANs) and allow backup images to be saved to a variety of different media. You are able to convert your image of your physical servers to virtual instances. You can convert your virtual images back down to physical instances.

Having multiple ways to recover your data allows you to create a suitable backup strategy for your environment and grow it into a disaster recovery plan.

You are able to have pre-exported recovery points which allow for minimal downtime in an event. This can be leveraged for maintenance, platform migrations, Dev testing and upgrades without having to rebuild servers from the bare bones up. That way, if something goes wrong with that change then you can bring the machine back to where it was before the change very quickly.

One of the main reasons that flat and file backups are chosen over full system or image based backups is because of target storage availability. Most customers don’t have the storage capacity available and can’t afford to put down a larger storage device. Because image-level backups use snapshots, all of the data including deleted files and empty disk blocks, are backed up. To reduce the amount of data stored: data deduplication. Arcserve UDP can achieve very impressive deduplication and compression ratios making it comparable to a file/flat backup strategy on target data consumption. For example, where I backed up full image backups of 179 Servers which totalled 7TB of raw data, it amounted to roughly 200GB held on disk after deduplication and compression.


File level backups basically come down to you selecting some files and folders that you want to back up and then where you want those file level backups to go.

Image/system state backups allow you to select an entire drive, partition, or entire machine which backs up the entire selection you have selected.

In my opinion there are more features on image based backups than on file. In addition, with UDP all you would need is one backup of your server and you can replicate, export, BMR etc.