A recent survey by Storage magazine finds that while the number of firms using a tiered storage model today is approximately the same as three years ago (55% vs. 51%), tiering practices have become more sophisticated with the addition of solid-state. In 2008, 16% of those surveyed said they used solid-state storage for tier 0; today, 42% have replaced spinning disks at the max tier. Still, most organizations top out at tier 1 using Fibre Channel (FC) arrays with 37% using 4 Gbps FC and 26% tiering on 8 Gbps FC.
When asked what the biggest pain point related to their tiered storage system was, classifying data so that it’s sent to the right tier (21%) was the No. 1 issue, followed by moving data between tiers (19%), establishing policies that determine data movement (18%), poor performance on lower tiers (16%) and keeping track of where data currently resides (14%).
Now seen as an enabler for the efficient use of solid-state and spinning disk technologies, storage tiering is also gaining ground as a storage efficiency technology. Storage magazine's recent "Storage Purchasing Intentions Survey" found 27% of data storage managers currently use automated storage tiering (AST) and another 32% will evaluate the technology.
To help you understand more about the tiered storage model and how it's being used in organizations today, we've collected some of our most popular content on tiering practices and the latest tiered storage technologies.
One of the biggest challenges users face when tiering their data storage is the idea of data classification and archiving. This podcast discusses the challenges associated with classifying data, and how you can automate the movement of data by deploying policies within applications or file system layers. We also discuss how data migration tools can help users when tiering storage, and when to move long-term archiving materials to a separate tier of storage.
Randy Kerns, senior strategist at analyst firm Evaluator Group, explains the differences between external tiering (tier 1, tier 2, tier 3 and sometimes even tier 4 storage) and "inside-the-box" storage tiering, which has gained more attention due to NAND flash solid-state drives (SSDs).
Manually tiering data can be a time-consuming process, so block-based storage systems with auto-tiering software that can move chunks of data to the correct tier could convince more users to make the change to solid-state drives (SSDs) and the boost in performance that comes from using SSDs. Senior writer Carol Sliwa speaks with analysts about the growing popularity of auto-tiering in data storage environments.
This podcast discusses the ins and outs of automated tiering. Learn the two things you need to do when starting an auto-tiering project, how to decide what information should continue to be tiered manually, concerns over assigning too much control to your auto-tiering environment and the rise of sub-LUN tiering.
Tiering is typically a block for block movement of a logical unit number (LUN) in its entirety. But not all data in a LUN is needed at the same rate; by instituting sub-LUN tiering, you can move chunks of data that are in high demand to faster storage tiers. This article discusses the number of storage tiers that need to be supported by auto-tiering software, how block size matters in automated storage tiering, and how long it will take your workloads to be analyzed and moved to the new tier of storage.
This article was previously published on SearchStorage.com.
This was first published in December 2011