In-Depth
Using Replication to Protect Remote Office Data
There are many advantages in selecting a replication solution over remote backup, from significant cost savings in hardware, software, and operations to better data protection, business continuity, and disaster recovery -- all in a single solution.
by Bennett Klein
Today, many businesses, government agencies, and higher education institutions face the daunting task of protecting the information stored at remote offices and branch offices typically referred to as ROBOs. Historically, some of these organizations have utilized a software backup solution. This strategy typically requires someone to perform the backup and manage the media (such as tape), whether it’s an IT administrator who travels to the remote office; a local reseller or service provider who helps the organization; or in the worse case, the organization relies on a local person (typically a secretary or clerk at the remote office).
Most organizations run backup on a daily or weekly basis depending on their recovery point objectives (RPO) -- the amount of data an organization is willing to risk in between periodic backups.
This strategy presents numerous challenges to these organizations. First, they have the cost of the backup software, a backup server, disk storage, and typically a tape drive or tape library and media. Next, they have the cost of a data center IT administrator traveling to the remote offices to run the backup and respond to failed backup jobs which could be caused by a hardware, software, or media failure.
Even if the IT staff uses a remote access solution to run the backup or restore job on a server in the remote office, there can still be issues if the necessary media is not installed in the drive, requiring someone to search for the correct tape and insert it into the drive. Outsourcing the backup job to a local reseller or service provider adds substantial cost.
Finally, for security and disaster recovery purposes, most organizations prefer to keep the backup tapes/media at a central location such as their data center or headquarters, requiring them to either hire a messenger service to pick up the backup tapes or, worse yet, request an employee at each remote office to bring the tapes into the headquarters following the tape backup. Additionally, organizations that perform disk-to-disk backup face the risk of damaged, lost, or stolen tapes.
Available Backup Options
One option is to use backup software to back up or copy data over the Wide Area Network (WAN) to the data center or headquarters. This provides centralized data protection and allows the backup and restore process to be performed by trained, experienced IT staff. However, many small and mid-sized companies often face the additional challenge that the large volume of data would take too long to transmit across the WAN, as many of them do not have a high-speed connection such as a T1 or T3 dedicated line between each remote office and the data center.
Even if backup across the WAN is feasible, it doesn’t address the recovery point objectives many organizations need to mitigate risks of accidental or malicious cyber attacks. Optionally, organizations may choose a cloud-based backup solution that leverages a third-party service provider’s data center, but that does not overcome these challenges.
A better option for some organizations might be a real-time WAN-optimized continuous data replication solution. This solution is available as host-based software, where the replication software is installed on each application server at the remote office, or as hardware-based solution as part of an appliance or storage area network (SAN) storage array. Hardware-based replication solutions are typically more expensive as the replication technology is secondary to the device, such as a SAN-based disk array. Software-based replication solutions present several benefits: they are typically less expensive, allow any-to-any replication to reduce storage costs, and are easier to implement and manage – especially for small and mid-sized organizations with limited IT staff.
All host-based software replication solutions typically work in a similar manner. First, each protected server -- typically referred to as the production server -- at the remote office is synchronized over the WAN or by some other method, with a similarly configured physical or virtualized server -- typically referred to as the failover or replica server -- at the data center or HQ.
Once the production server and failover servers are synchronized, only changes in the data on the production server are transmitted in real time across the WAN, enabling efficient transmission over a limited bandwidth connection between the remote offices and data center. This continuous replication keeps the production server and failover server synchronized -- using a near-mirroring-like method -- but performed asynchronously to keep WAN bandwidth requirements to a minimum and to deliver better performance.
More comprehensive host-based software replication solutions include other capabilities such as continuous data protection (CDP) where the IT administrator can quickly and easily rewind the data on the failover server back to a “known good point in time” in case of accidental or malicious data loss or damage on the production server at the branch office. Then the recovered files or database can be restored back to the production server.
An added benefit to data replication over a remote backup solution is the ability to add a server monitoring and automated failover option, offered by some host-based replication software vendors, to the core replication solution. This provides a high-availability (HA) environment for critical and near-critical applications and data -- helping each branch ensure data protection, business continuity, and disaster recovery (DR) -- all in a single solution. For the best risk mitigation, high-availability replication solutions also include automated, non-disruptive DR testing that ensures the failover environment at the data center is ready whenever the branch requires failover.
Once the data from all branches is replicated to the data center, the IT staff can run the backup off the local failover server for historical data protection, data archiving, and regulatory compliance purposes. In this scenario, running backup does not affect the production servers at the branches because the backup is run off the failover servers instead, eliminating any impact to the remote office productivity and also helping overcome backup window constraints most organizations face today. This solution also eliminates several issues related to remote tape backup such as the need to send IT staff out in the field (or contract with a local service provider) and the cost and maintenance associated with backup servers, tape drives, libraries, and media at each branch. It also ensures centralized, secure data protection at the data center or headquarters, where it belongs.
There are many advantages in selecting a replication solution over remote backup, from significant cost savings in hardware, software, and operations to better data protection, business continuity, and disaster recovery -- all in a single solution.
Bennett Klein is a senior director of product marketing for CA XOsoft Replication software. He has more than 20 years of experience in marketing of data protection, storage, and high-availability software for enterprises of all sizes around the world. You can contact the author at Bennett.klein@ca.com.