With Backup Express 3.0 with XRS, Syncsort has provided a backup solution that uses near CDP to offer a completely new restore experience for customers. The solution provides the ability for recovery in place, server re-imaging and thin cloning.

The key to Backup Express 3.0 is the integration of all the parts offering the ability to replace point solutions in the data center. One of those point solutions is snapshots. They have grown from a good idea to where, in the data center, we are seeing them either not be used or become a major problem.

To make sure we are all on the same page, let's explain what I mean by snapshots. The file or volume being 'snapped' is comprised of blocks of data.  These blocks are organized by an index of pointers or links to the actual blocks on disk. When an application requests data, it does not try to find the blocks itself, it is routed to the index for the location. When a snapshot is "taken" you are actually only copying this primary index. This is very small and takes less than a few instants, if that, to complete. The blocks that are referenced by that snapshot are then put into a read-only mode and can't be changed or deleted.

Typically upon the initial snapshot almost no additional storage is consumed. As the data is altered throughout the day, new blocks representing data appended to the original file or volume and blocks that have been changed are written to a new area of disk.  Only the primary or active index is updated with those block locations. This is also where data growth begins. The snapshot is not updated and still references the static read-only blocks, allowing you to view that data at a point in time.

The first key issue is the way data is stored in a real world data center. Data is not simply stored on one SAN. Data is still stored locally in many cases; there are multiple vendors' SANs in most data centers and typically a NAS with its own snapshot method. The challenge is that each of these storage areas requires a unique snapshot and process. It is not uncommon now for there to be a dozen snapshot processes running in a data center and all those have to be checked and managed.

What is needed is the ability to bring that under a single umbrella, so that a single process can provide snapshot-like functionality across every platform. The problem with this has been that the last thing IT managers want to do is install yet another agent on a server. The advantage that Backup Express XRS has is that the snapshot agent comes bundled in with the backup client. You already have a backup agent installed on most servers, and by installing the Backup Express agent you get backups and the ability to do snapshots in one move.

With the agent in place, you can run block level incremental backups every hour if you choose. You leverage the agents that are available as part of the backup process to make sure you get clean copies of databases and email servers. At the beginning of the backup process the information already on the XRS server from the prior backups is snapshotted, then only the new or changed blocks of data are sent to the XRS server. This gives you the ability to roll back to an older backup for a point in time for recovery or pull from the latest backup, all done with minimal impact to the servers being backed up.

These snapshots are now Enterprise-wide, not limited to data on the SAN. They are also controlled from a single application, not from a different application for every platform. Some backup applications will "manage" other snapshot's processes from their GUI, for example triggering a snapshot on vendor A's SAN. While an improvement, you have to hope that your backup application will support your SAN or NAS and that it does an effective job managing it. Additionally, will it manage the steps required to get the databases and email servers into a backup mode? Backup Express replaces and improves the entire process.

As a result, you have off-loaded the snapshot process from the original server and the storage. Most SAN storage for example is extremely limited as to the amount of snapshots that can be maintained and as a result, users typically keep very few snapshots and tend to only perform them a few times a day. The performance impact to the storage area of keeping even a handful of snapshots maintained is severe. In many cases three or four snapshots can impact production performance by as much as 50%. In the case of Backup Express, since the XRS server maintains all the snapshots, there is no impact to server or storage performance. All the overhead is offloaded to the XRS server and since the XRS server is essentially dedicated to the task of maintaining snapshots, the amount you can maintain is significantly higher.

Most snapshot methods are "Copy on Write" meaning that as a block changes the old block is written to a snapshot area. That snapshot must be stored on the same disk array with the same RAID format. This is an expensive way to maintain old versions of data. With Backup Express, the snapshots are maintained on the array attached to the XRS server. That storage attached to the XRS server can be from virtually any storage manufacturer. For example, it can be an inexpensive SATA based array. This will drive down the amount of primary storage you need and allow you to use primary storage for what it is intended, for delivery of active data sets, not storing data for recovery.

In addition, traditional snapshots are totally dependent on the primary storage to remain valid. If you have a failure in your primary storage, all your snapshots fail right along with it. There is no way to recover from the snapshot if your primary array has failed. Traditional snapshots are only a viable solution to corruption of a single file or groups of files in a database. They are not a solution in the event of a storage failure. You are using primary storage real estate as a recovery option for only one data loss scenario. With Backup Express XRS, you can use the snapshots to recover from the corruption of a single file or group of files like a snapshot, but you can also use it to recover to a totally different location in the case of a complete storage hardware failure providing for server re-imaging.

Additionally, because of XRS's unique capability, you can recover in place. With proper planning, you can iSCSI mount the recovery data and be back in production in a matter of minutes, without having to move data. In the case of a server failure, a recovery CD can be made and restoration of the server itself can begin within a few minutes. In a typical SAN you are booting off of local storage and even if you were snapshotting that local storage, in the event of a server failure those snapshots went down with the server. Since XRS makes it easy to snapshot across storage platforms and offloads recovery data to a separate server and storage device, you are able to recover from almost any failure scenario.

Unfortunately, you can't afford the expense of keeping snapshots online forever, especially with traditional snapshots that store their snapshot information on the same primary storage as the storage being snapshotted. With traditional snapshots, moving that data to secondary or longer-term storage can be a challenge. Backup Express is first and foremost a backup application; the move from the snapshot disk to a more suitable long-term storage device is essentially built in and does not require a separate process. With traditional snapshots and backup applications, you have this very complicated dance to time the snapshot process and the backup process. This process usually requires special scripting and is very sensitive to changes in the environment.

Secondary storage can of course be tape or a disk target that can, for example, perform data de-duplication and replication. This allows you to keep the XRS target disk relatively small by pushing the aging data sets to a secondary and less expensive target.

With Syncsort Backup Express 3.0 and the new XRS features, we think that Syncsort has made a strong case for consideration not only as a backup application but also as a complete data center protection tool.