Arshad Munir Sharif, Storage Leader, IBM ASEAN
DataStorageAsean: Archive is an old technology, why is it becoming hot once more?
Arshad Munir Sharif: Archiving technologies have been around for a long time. The technologies associated with lowest cost archiving were originally used because active systems technologies could not keep pace with the volumes of data being created. But with the exponential growth in data spurred by connectivity, archive technologies are here to stay. Market research has shown that the volume of data is expected to reach 40 zettabytes by 2020. And storing aged data is a necessity although it impacts enterprise IT budgets. IT professionals need to review budgets to manage aged data carefully as studies show that 80% of data in storage is barely used 90 days after it is created.
Archive technologies offer efficient data management and long term retention for inactive data or low activity data, replacing tier 2 or 3 disk with tape libraries for economic tiered storage and lowers total cost of operation. IBM uses automatic policy-based storage tiering that moves data from flash through disk to tape and cloud tiers, allocating key data to highest performing tiers and lowering costs by moving “cold” data to lower cost tiers.
DataStorageAsean: Is archive just about file data or is it much more than that?
Arshad Munir Sharif: In the past, organizations were focused solely on the structured data in databases and data warehouses. Now, organizations want to mine unstructured data from social media and mobile applications. Archiving is used in both structured (Block) and unstructured data (File). No matter what type of data is generated, the need to retain and repurpose the data for analytics or to retain it for legal holds requires tiered storage management including an archive tier. In a perfect world, data would move to the fastest storage for analysis, and then retreat to lower-cost storage when not in use. It is critical to use policy engines and analytics-driven data management to keep data in the right place automatically, based on usage, and move data between storage systems without disrupting users or applications. This allows clients to run their big data and analytics projects and environments with much-faster performance at a much-lower total cost.
DataStorageAsean: How about archive medium? Does tape still have a place or is it all about cloud?
Arshad Munir Sharif: For large archives where data is consolidated or where customers have large amounts of data, tape is still the choice today and for the future. That is because it offers the lowest total cost of ownership for long-term retention/archiving of data. The cost of IBM tape / IBM spectrum archive solution is only 0.2 cents per GB per month which is a 300% savings over the average cost of cloud service provider.
The figures speak for themselves: Tape is not going away. Over 24 exabytes of tape storage were shipped in 2014, the highest in its 63 year history.
DataStorageAsean: How about Archive as service - is that a viable model?
Arshad Munir Sharif: As stated earlier, the explosion of data is disrupting traditional storage models. As long as businesses need to keep data for the long term and make fast data driven decisions, they will need to look for innovative ways to derive value from the tremendous volumes of data available while achieving superior data economics. To accomplish this goal businesses will need to adopt newer storage architectures and delivery models such as Software-defined storage (SDS) and cloud. IDC predicts that by 2019, 70% of storage providers will offer software-only storage. When you compare this to IBM, we are already there and in production.
IBM Spectrum Storage range of solutions optimizes clients’ costs, performance and security. Archive as a service, Managed Storage, Cloud Archive, no matter what the name, are viable models for archive/long term storage of data. In order to be successful, "archive as a service" models will have to adopt technologies that offer the ability to tier data to very low cost storage. The storage will have to be based on proven technology that needs very little management and provides very high reliability for the stored data, all of which are fundamental design principles of tape based solutions.
DataStorageAsean: Does the ASEAN region have any uniqueness which lends itself to needing Archive technology?
Arshad Munir Sharif: According to IDC, ASEAN consumers generate a higher percentage of data related to unstructured "entertainment" based data. This type of data goes cold very quickly. Since the data goes cold quickly, it is well positioned to be moved to an archive tier of storage much sooner than other data. Moving the data to an archive tier of storage sooner means the impact to the storage budget is lower sooner.
A very interesting example is the solution IBM implemented at the John F. Kennedy Center for the Performing Arts. The Kennedy Center captures about 5 TB/day of high quality video and audio that is freely available online. The mission was to preserve all performances by digitizing, protecting and providing rapid access to them with high availability all streamed live on line and also archived and available for users on demand. The IBM Storage solution helped them to achieve a 90% reduction in storage costs per terabyte over three years using tape, reduced administration workload, and cut operational costs through automation.
DataStorageAsean: What is unique about your own offering?
Arshad Munir Sharif: IBM Mis the worldwide tape market share leader offering a full range of tape storage solutions including drives, autoloaders, libraries, virtual tape systems and Spectrum Archive software.
IBM Spectrum Archive is part of our IBM Spectrum Storage family which is the #1 provider of SDS platforms offering a software defined implementation of end to end solutions for data management and optimization. The Enterprise Edition can run any application designed for disk files on tape and can play a major role in reducing the cost of storage for data that does not need the access performance of primary disk.
For instance, the IBM Spectrum Scale and Spectrum Archive will quickly store and manage unstructured data with Information Lifecycle Management that enables administrators to set policies to manage data based on the value of the data. This means data is stored in the lowest cost tier based on last access and that data moves to the archive tier automatically.
This is well demonstrated with EVRYs seismic data repository that ingests and stores 4 to 5 TBs of raw data each day and then must store 10 to 100 times more when that data is processed into different formats. Infrastructure is easily scaled with Spectrum Scale and Spectrum Archive while data remains 100% available.