How Image Based Data Protection Works
The end goal is to be able to produce an image of the protected volume(s) as of any snapshot point. The Reliable Data Protection solution achieves this goal through the use of two components on the local network. The first component is the core server. The core server holds the images of the protected volume(s) as well as the ensuing snapshots. This server is also capable of taking the image and snapshot data it holds and producing a runnable server image of any snapshot point. The core server can be on a dedicated server or it can co-exist with another physical server. The second component is the agent, which we will discuss in detail below;
1. First, a very thin agent is installed on every machine to be protected. This agent positions itself between the file system to be protected and the drivers for the disk or RAID hardware. This agent is responsible for sending the initial image to the core server. It also keeps track of the volume blocks that change between snapshots.
2. Step One; The client side agent starts off the process by taking an entire block level image of the protected server and sends it to the core server. The core server takes that image, performs data de-duplication and compression, and then stores it.
3. Step Two; The agent “watches” the protected volume and keeps track of all blocks that have changed since the initial image or last snapshot. When the next snapshot interval comes due, the agent engages VSS to settle the volume and allow it to take copies of all blocks that have changed since the last snapshot.
4. Step Three; The agent sends a current copy of all blocks that have changed since the last snapshot to the core server. The core server does several things at this point. It verifies that it indeed did get the correct quantity of blocks and they are good. To the extent the protected volume houses an Exchange or SQL server, it mounts the Exchange or SQL instance to ensure it is not beginning to corrupt. Finally, it stores the de-duplicates, compresses and stores the snapshot.
5. Step Four and Five; A customer may choose data center based storage for disaster recovery purposes. In that case, the initial image is sent over an encrypted channel to the data center where it is received by a peer core server. This initial image can also be sent via removable storage, such as USB drive, to the data center in a process known as seeding.
6. Step Six; The snapshots are sent to the data center for disaster recovery purposes. A unique feature of the Reliable Data Protection Image service is that the recovery points in the data center do not need to mirror the local recover points. For example, you may have 15 minute snapshots locally, and 6 hour snapshots in the data center.