Download the authoritative guide: Enterprise Data Storage 2018: Optimizing Your Storage Infrastructure
If you are like me, by now you are growing tired of reading about predictions and speculation of what will be hot in IT and storage for 2007. Rather than take the usual ivory tower route of predicting what will be hot or popular among the latest trends and what you should look at in 2007, I'm going to share with you what I will be looking at and considering for my own "virtual data center" to address various challenges that fall under various mega-trend categories.
In general, a mega-trend is a widespread major trend that can be broken down into sub-trends. Two examples of mega-trends include conservation and data protection. Conservation can be broken down further to include consolidation of resources using virtualization techniques and technologies to reduce electrical power, cooling and floor space requirements while improving asset utilization, service delivery and ease of management. Data protection can be broken down into backup and recovery, disaster recovery and business continuity, security and data preservation (archiving) for compliance and other purposes.
After a disappointing early trial with CDP technology, I want to take another look and keep an eye on how CDP techniques and technologies converge with traditional data protection technologies, including backup, snapshots and remote replication. My primary goal for CDP is to be able to reduce the amount of time backup takes to scan and determine which files to back up for faster backups and less I/O wasted on overhead. Of course, the benefit of being able to have variable dynamic recovery point objectives will be a nice benefit.
For scalability of performance, capacity, ease of upgrade and availability, I am keeping an eye on clustered storage solutions, including iSCSI and Fibre Channel block-based, NAS-based and file system-based solutions. While my applications are not high performance compute-intensive, the ability to start with a small block- or file-based clustered storage solution and scale incrementally is very appealing to support data migration needs and growth.
From a data security standpoint, both logical and physical security needs to be addressed, including securing facilities and technologies. Protection of data means ensuring that desktops and laptops are encrypted along with removable hard disk drives (RHDD), magnetic tape, data over networks and data at rest on storage systems.
Adding more storage is a given. E-mail archiving and data preservation and improved availability and survivability using a secondary site will be important issues. Data migration technologies are also important, including transparent movement of homogenous and heterogeneous storage using host-based, appliance-based network solutions or storage system-based tools combined with secure shredding and erase technologies.
For remote and branch offices (ROBO), consolidate where possible, and where it's not possible, look to improve resiliency, including network-based backup, removable media, fire-resistant and water-resistant media and other tools. Wide area data services (WADS), also known as WAFS or WADM tools, and remote or network-based backup tools and managed service provider backup have the potential for addressing ROBO data protection. Another technology for ROBO environments is hybrid, unified storage solutions that support remote replication to an alternate site or service provider.
Facilities concerns include reducing the amount of electrical power required to stay within the available power budget as well as to reduce costs. Reducing power means addressing cooling and ventilation costs as well as costs for running equipment. I'm also looking into enhancing standby and backup power capacities to address outages due to storms, disasters or rolling power blackouts. Reducing power ties into the mega trend of conserving, since it involves improving utilization via virtualization and performance acceleration to reduce component counts, as well as to replace dated, less efficient technologies.
As a means of reducing costs, the buzz around SRM has dwindled, and the new buzz is around infrastructure resource management (IRM), performance and capacity planning tools that let you optimize storage capacity utilization. However, I'm concerned about over-optimizing storage capacity at the expense of service delivery, performance and availability.
Consequently, I will be looking at tools that provide a balanced picture of improved asset utilization while improving performance and availability. Leveraging automation through rules based-policy managers for tasks such as data migration, archiving and automatic deletion of expired data may help address routine management tasks. Other tools I see as helpful are those that perform cross technology domains (servers, storage, applications, networks and so forth), event correlation and root-cause analysis for problem determination and "what if" analysis.
While not a technology, staff and personnel development, including continuing education and training, is important to keep up on all of the buzzwords, new technologies and techniques. Leveraging webinars, conferences and other sources of information, including networking with peers, is important.
Buzzwords and hype that I am anticipating hearing more about in 2007 include configuration management databases (CMDB), ultra-large scalable (ULS) storage systems, hybrid and unified storage with federated management, among others. Holographic storage is becoming reality, with initial beta units now in the field and production systems anticipated later in the year. Holographic storage has good potential for future archiving and data preservation applications once the technology matures and pricing comes down with volume, so for now, it is on the watch list.
In general, I'm looking for tools and technologies that are easy to acquire, install and use on an ongoing basis and provide good effective value, and in some cases, that can fund themselves via productivity or some other benefit. Regardless of whether yours is a virtual data center or a real data center moving to a virtualization-enabled IT infrastructure, there are plenty of challenges and needs to be addressed using various techniques and technologies. Aligning the right technology with the task at hand and making sure those solutions can scale with stability and not add complexity or management burden remains the objective. Rest assured, 2007 looks to be yet another interesting year for data storage and IT infrastructure topics.
Greg Schulz is founder and senior analyst of the StorageIO group and author of "Resilient Storage Networks" (Elsevier).