The Case for Unifed Data Protection - Page 2


Want the latest storage insights?

Download the authoritative guide: Enterprise Data Storage 2018: Optimizing Your Storage Infrastructure

Share it on Twitter  
Share it on Facebook  
Share it on Google+
Share it on Linked in  

Architectural Requirements for Next-Gen Backup Platforms

Let’s take a look at the requirements for a next-gen backup platform.

New architecture

Detailing the Requirements

  • Near-Continuous Data Protection. Traditional backup depends on making copies. A lot of copies. Most admins schedule incremental backup to control backup size and length, but they must run full backups at least once a week and once to twice a day on Tier 1 applications. In contrast, new data protection architecture takes a baseline image and combines it with CBT for incremental-forever backup. In addition, the changed blocks are deduplicated across the entire backup infrastructure for real efficiency and capacity savings.
  • Extreme Recoverability. Recoverability depends on two factors: 1) how trustworthy is backup data integrity and 2) how fast can the system recover applications and data to a working state? In answer to the first question, backup verification is immensely important as recovery speed means nothing if the restored data is compromised. Verification should test that a backup completed correctly, even when backup is running every five minutes. Specific application verification is also a big plus in next-gen platforms. Application testing gives IT and application admins high confidence in backup and restore integrity for critical systems. As for instantaneous recovery, the backup system should be able to priority-queue backup data for immediate restore. The system should also be capable of quickly restoring applications to working order while full data restoration occurs in the background. This capability can save hours to days of downtime with applications, a huge benefit to both IT and users.
  • Efficient Cross-Domain Support. A number of backup applications support both physical and virtual operations, but the operations and management interfaces are different. This introduces backup system complexity and yet more management overhead. New architectures provide backup and recovery for physical and virtual systems, applications and operating systems—and they do it using shared backup features and a common management interface. Leveraging shared features such as global dedupe unifies backup operations across domains for greatly increased efficiency. An easy-to-use common management console is highly efficient and diminishes management overhead.
  • Cost-Effective Operations. Next-gen platforms save money on backup resources by using CBT/incremental-forever to accelerate backup times, shrink capacity requirements and greatly accelerate restores. They should offer built-in fast and flexible replication with strong encryption options like AES-256. Modern platforms efficiently dedupe and compress data for fast data movement without the cost of additional equipment. Scalability is also important for optimizing backup and recovery costs across multiple applications, networks and operating systems. Petabyte storage levels along with deep data compression will yield high scalability, and connecting platform systems to one another will further scale storage load balances, common technology features and centralized management.

What about agents?

For years, debates have raged over the use of agents in backup software. Agentless solutions do have benefits, such as no extra load on the protected server and no need for agent management features.

However, agent-based solutions have strong advantages as well. Agents provide application awareness such as the ability to gather metadata about the application including application versions and components, configurations, and the ability to truncate database logs.

Intelligent modern agents act in concert with the host to quickly carry out backup and recovery operations directly on the protected machines.

Furthermore, even an agentless system has to assign resources to backup and recovery operations. In their case, this intensive processing occurs in the host, which can require far more resources than distributing operations among agents.

Our Conclusions

Near-continuous backup, near-immediate recovery and broad domain support are not easy to do. When you add in cost-effectiveness, you have a real challenge on your hands.

Vendors are certainly trying because there are rich rewards for success in the data center. Many backup vendors offer scalable solutions with centralized management. Some vendors concentrate exclusively on virtual networks, but the market leaders are attempting to protect a variety of environments in the backup infrastructure.

Running multiple backup tools can be complex and costly, and extreme data growth and virtualized networks are worsening the problem. We strongly suggest that IT consider platform backup technology that unifies backup needs across the data center, and that is founded upon scalability, ease-of-use and exceptional efficiency. The more backup operations that IT brings under the platform’s control, the greater the company’s return on investment.

Christine Taylor is an analyst specializing in data protection and information governance at Taneja Group.

Submit a Comment


People are discussing this article with 0 comment(s)