dcsimg

Guide to Tiered Storage

SHARE
Share it on Twitter  
Share it on Facebook  
Share it on Linked in  
Email  

Tiered storage is a storage method which involves storing data on a range of different storage media with different characteristics, such as performance, cost, and capacity.

The different storage media are organized into a hierarchy, where the highest performance storage media is deemed to be Tier 0 or Tier 1, followed by Tier 2, Tier 3, and so on.

Tier 0 or Tier 1 is usually made up of flash or 3D XPoint-based solid state drives, while successive tiered storage levels may involve high performance fibre channel or SAS drives (or RAID arrays containing them), lower performance SATA drives, optical disks, tape storage systems, and cloud-based nearline or offline storage systems.

A 3 tiered storage model using disks and tape to provide tier 1, tier 2 and tier 3 storage was perhaps the most popular tiered storage model before SSDs and cloud storage became commonplace.

But now it is not unusual to use a tiered storage model that involves five or more tiered storage levels. Each one is subtly nuanced to produce a different combination of the three key storage attributes: cost, performance, and capacity.

The Objective of Tiered Storage

If expense were no object, businesses would use SSDs for all their storage requirements because they offer very high performance and reliability.

But in the real world storage costs are very important, as IT departments seek to live within their budgets and organizations as a whole seek to minimize costs and maximize efficiency. And unfortunately SSD storage is expensive compared to hard disk drives, and far more expensive than tape storage.

That means that SSDs need to be used judiciously, and only for storing data that is used  by systems that require very high performance. Data that is less important can be stored on lower cost, lower performance systems such as hard drives, and data that is rarely accessed, or is only retained for compliance purposes, can be relegated to very low cost offline storage systems

So tiered storage is not driven by the needs of the IT department for operational reasons, but purely for financial reasons. The objective of a tiered storage system then  is to minimize storage costs by using the least costly storage option possible which provides the minimum required performance.

How Does Tiered Storage Work?

A tiered storage system with just two storage tiers offers storage administrators a very limited choice over where a given piece of data should be stored. If a certain level of storage performance is required for that data then it will be stored in Tier 2 if that offers the required level of performance, or, if it does not, then it will be stored in the more expensive Tier 1.

A tiered storage system with three storage tiers is likely to be more efficient, because that same piece of data can be stored in Tier 3 if that offers the required level of performance, or if not then it can be stored in Tier 2 if that offers the required level of performance, and only when both Tier 2 and Tier 3 fail to offer the required level of performance does it need to be stored in the most expensive Tier 1 storage.

In fact the more storage tiers there are available, the more efficiently a piece of data can be stored in an appropriate storage tier that meets its performance needs, rather than one that which offers unnecessarily high performance at unnecessarily high cost. New tiers can be created from a small set of different storage media types using techniques such as creating RAID arrays or short stroking disks. Both of these techniques, for example, create higher performance storage at the cost of lower storage efficiency (and thus, ultimately, higher financial cost).

The biggest challenges presented by a tiered storage system are classifying data into multiple classes, deciding which storage tier best suits a given class of data, and reclassifying data on a regular basis as it ages.

The key point here is that any given piece of data's storage requirements may change over time, so rather than storing data in a tiered storage system and forgetting about it, the data has to be monitored and moved to a lower (and lower cost) storage tier as soon as it no longer requires the performance of the storage tier that it currently finds itself in.

Typical data classes for tiered storage include:

  • Mission critical data. This class of data that always needs to be stored in the highest level of tiered storage because it is needed to support high-speed applications – perhaps supporting customer transactions. Delays in accessing data will cause the organization to lose business or otherwise negatively impact profitability. Performance is all-important.
  • Hot data. This class of data needs a relatively high level of tiered storage  because it is in constant use in applications such as CRM, ERP or even email, and needed for the day-to-day running of a business. Performance is important at this layer of tiered storage, but cost is also a consideration.
  • Warm data. This class includes older data such as emails that are more than a few days old, or data on completed transactions. This type of data will be accessed relatively infrequently but still needs to be readily accessible when required. The most important consideration at this layer of tiered storage is cost, but subject to a minimum performance threshold.
  • Cold data. This class of data may never be accessed again, but it needs to be archived and retained to comply with regulatory or other legal requirements, or simply because it may have some value at some unspecified time in the future – perhaps for big data analysis. Cold data is ideally suited to the lowest layer of tiered storage where access times of minutes or hours are acceptable, and low cost is the overriding consideration.

Apart from the very smallest businesses, it is clear that a manual approach to storage tiering is likely to be too time consuming and cumbersome to be successful. For that reason, most tiered storage systems rely on automated storage tiering using a tier management system which monitors data throughout its lifecycle, and moves it down the storage tiers automatically as it cools.

This software may be supplied with, and operate on, a single storage system, or it might be a standalone solution which works right across an organization' storage infrastructure, and in some cases out into the cloud as well via a cloud storage gateway.

Tiered Storage Types

What does "Tier 1" mean?   

In some organizations which have used storage tiering for some time, new, faster storage technology than used in their existing Tier 1 has become available. For that reason, this technology is adopted as "Tier 0" storage. But for the purposes of this article Tier 1 will be assumed to mean the highest performance storage tier.

Tier 1

This storage tier is designed to store data which is highly volatile and time-sensitive and which needs to be available in the shortest possible time. Typically used in financial trading environments or other business areas where far larger amounts of money than the cost of the storage can be made or lost in a fraction of a second, only the fastest possible storage is sufficient.  For that reason very fast solid state storage media configured for the highest possible performance, regardless of overall storage efficiency, is usually required for Tier 1 storage.

Tier 2

This storage tier is used to store transactional data to support high-performance applications, customer-facing systems such as retail applications, and other systems where anything more than extremely short delays are unacceptable. Since the performance levels required are not quite as high as for Tier 1 purposes, lower cost and more storage efficient solid state storage systems are typically used to provide a Tier 2 storage solution.

Tier 3

The third storage tier is used to store "hot" data such as CRM and ERP data, and even recent emails, which needs to be accessed frequently without undue delay. That means that it needs to be stored on media such as medium to high performance hard drives which are relatively low cost compared to solid state storage media. In many organizations there will be substantially more Tier 3 data than Tier 2 or Tier 1, so high capacity hard drives which offer a low cost per Gigabyte stored provide an ideal solution.

Tier 4

This storage tier is used to for "warm" data such as data from recently completed transactions, emails which are more than a few days old, and other data which will be accessed infrequently.  It can also be used to store data which may be used for business information and data visualization system, or financial data which will be needed for creating monthly, quarterly or annual reports.  but nevertheless needs to be accessible relatively quickly whenever required. Tier 4 storage requirements can be very large, and since cost is a major consideration a typical solution for Tier 4 storage is high capacity hard disk drive storage. Rather than high performance SAS disks or RAID arrays configured for performance, Tier 4 storage is more likely to use low cost, relatively low performance SATA drives.

Tier 5

The lowest storage tier is typically used for archiving "cold" data which may never be accessed again, but which still retains some value (perhaps for data mining in the future) and which is therefore worth retaining. It  is also used to store data which must be retained for compliance purposes, and which only needs to be accessible within days or weeks rather than minutes or seconds.

Optical media or tape storage systems which offer storage at a very low cost per Gigabyte, but with very low performance, are typically used for this storage tier. In the last few years cloud-based storage such as Amazon's Glacier storage service has become popular for companies who wish to offload large amounts of data for low cost storage with the possibility or retrieving it relatively easily in a matter of hours if ever required.

Key Advantages of Tiered Storage

Tiered storage brings a number of benefits, the most important of which are:

  • Reduced storage costs: by storing  each data class on the lowest cost storage that offers at least the minimum performance that it requires, companies can avoid paying for high performance that they do not require.  This reduction in storage costs is the primary driver for adopting a tiered storage system.
  • Higher storage efficiency: High performance disk storage systems which rely on RAID or short stroking are storage inefficient because the amount of data that can be stored on these systems is less – much less in some cases – than the total amount of storage capacity provided. Since tiered storage systems can relieve the demand on RAID and short-stroked storage (by enabling data that does not require such high performance to be moved to lower storage tiers), it follows that storage efficiency will go up.
  • Ability to reuse older storage equipment: Storage tiering can provide a new lease of life for storage systems which would otherwise need to be replaced because they no longer offer sufficient performance for upper tier data uses. Instead they can be used to provide lower tier storage.

NewsletterSTORAGE WEEKLY

Want the latest storage insights?