Originally posted by: Andre Gaschler
By Andre Gaschler, Nils Haustein, Jörg Walter and Erik Franz
I recently worked with my colleagues on an enterprise backup proposal where we designed a Software Defined Data Protection architecture. After coming back from the TSM Symposium 2015 where we learned about the new and great enhancements of the next generation data deduplication in IBM Spectrum Protect, which IBM announced about two weeks ago, I’d like to share our Software Defined Data Protection architecture with you.
The Situation we are facing
We all noticed that organizations are storing and using more data than ever before. Data volumes are growing exponentially. Government regulations are forcing organizations to retain more data for longer periods of time. At the same time competitive pressures increase the need to seek for more efficient methods to economically manage backup data. Therefore the necessary question arises, how to effectively manage enterprise backup environments while maintaining flexibility and scalability.
Furthermore, as data volumes grow, we see that storage administrators are challenged to complete backup operations within established backup windows. More data inside the backup system means that it will take longer to recover when something goes wrong and we all know this is increasing the operational and business risks.
One solution to all this data growth in the past has been to simply buy more storage. This sounds reasonable at first sight, because we have seen that the cost of the storage itself has been decreasing over time. On the other hand, the cost of housing, powering, cooling and managing all these devices are exploding. Environments become more complex as the number of devices is growing. We have noticed a trend a couple of years ago to make the storage more efficient by adding data deduplication and/or compression functionalities. However, while storage and backup teams spent lots of efforts to make the backup storage more efficient, it was often forgotten to ask the following simple questions:
- Am I storing the right data?
- How can I manage my backup data more efficiently instead of trying to store it more efficiently?
How we cope with these challenges and how we solve them
In our Software Defined Data Protection (SDDP) solutions we are addressing those challenges from an end-to-end perspective. The two key parts in our SDDP architecture are shown in the picture below:
- A backup concept for the backup data management
- A backup software for the backup data optimization plus an intelligent and transparent backup data placement
The backup concept is represented by the Modular Data Retention Concept (MDRC) and as backup software foundation for the SDDP solutions we are using IBM Spectrum Protect, formerly known as IBM Tivoli Storage Manager (TSM). The MDRC and Spectrum Protect layer are supported by our Software Defined Storage Portfolio where IBM Spectrum Scale and IBM Spectrum Virtualize as storage software is utilizing one our comprehensive and powerful storage hardware, like for example IBM Storwize.
Where backup architectures in the past required dedicated hardware solutions for data protection environments, our Software Defined Data Protection approach is lifting these dependencies by leveraging for example backup data optimization features included in the IBM Spectrum Protect software.
As you can read in the latest announcement we achieved performance improvements for scalability and efficiency of software defined deduplication up to ten times in capacity per Spectrum Protect server and up to four times in backup throughput using highly efficient in-line server-side deduplication technology.
In addition, the Software Defined Data Protection approach stands for smarter backup and recovery management by using our Modular Data Retention Concept (MDRC) to manage and structure your backup data already from the beginning in three layers:
- The Data Type Layer where we define the backup client server or application type and the required backup client software
- The Backup Parameter Layer where we define the appropriated backup retentions and backup schedules for your data type
- The Storage Infrastructure Layer where we choose the best storage option to finally store your backup data on
We classify each of your backup data types according to their backup & restore requirements and group them into data modules. These modules use then the backup optimization techniques from Spectrum Protect for example the industry leading incremental for-ever backup and the backup data deduplication functionality. Finally we store the backup data optimized using Spectrum Protect and its comprehensive storage hierarchy management, which allows transparent data placement on different storage layers, including local storage devices and cloud storage.
Smarter backup and recovery management can be built with Software Defined Data Protection solutions. The backup data growth and the challenges which will come along with that requires a more flexible and dynamic way of backup data management. The Modular Data Retention Concept (MDRC) and Spectrum Protect are already a perfect match. Knowing this, together with our Software Defined Storage Portfolio including IBM Spectrum Scale and IBM Spectrum Virtualize and our powerful storage hardware, for example IBM Storwize, IBM Flash or IBM Tape storage we are able to build and deliver end-to-end solutions which are designed for the future of backup.
The strategy is clear: Start managing backup data in an automated, policy driven and transparent way by using our Software Defined Data Protection solutions, instead of just putting it into endlessly growing storage silos.