News
6/25/2010
02:42 PM
George Crump
George Crump
Commentary
Connect Directly
RSS
E-Mail
50%
50%

The Types Of SSD Cache

In our last entry we discussed the value of using solid state disk (SSD) as a cache, which provides a simpler on-ramp to the accelerated world of SSD. With SSD cache there are no or limited changes needed to applications and using SSD as a cache does not require a large capacity investment in the more premium priced technology.

In our last entry we discussed the value of using solid state disk (SSD) as a cache, which provides a simpler on-ramp to the accelerated world of SSD. With SSD cache there are no or limited changes needed to applications and using SSD as a cache does not require a large capacity investment in the more premium priced technology.SSD as cache can benefit businesses of all sizes, whether you are a small business trying to get better performance out of your Exchange environment or a large one trying to optimize your Oracle environment. They also can be an effective money saver as well by delivering improved performance without having to add more or faster drives. Suppliers have been quick to jump on this market opportunity and there are now several cache solutions for direct attached storage, network attached storage (NAS) and storage area networks (SAN).

The simplest case is to use an internal SSD either as a drive or a PCIe card inside of a server. Several RAID controller manufacturers have added the ability to support an SSD drive to act as a cache in front of the drives attached to the RAID controller. For the cost of a single SSD drive, or two for redundancy, we've seen dramatic improvements in performance. The SAN can also benefit from cache and the solution does not necessarily have to come from your storage vendor. Several storage virtualization products have the capability to set aside several SSD drives as caching area for reads and in some cases even writes, to existing storage arrays. There are also a few products that are essentially caching switches, and don't require a storage virtualization engine. As is the case with the server example, this can greatly improve performance.

The above two examples cover the block world but NAS systems do not need to be left out. There are several devices that can be placed in front of NAS systems to cache CIFS and NFS traffic as it goes across the network. While many assume that the network is the bottleneck in high performance NAS environments, the thrashing of disk I/O is sometimes the root of the problem; caching should help alleviate that.

Most cache systems are temporal in nature. Meaning that they only hold a unique copy of data for a very short period of time, typically measured in seconds. In fact read only caches never hold a unique copy of data. We are seeing caching devices that are less temporal in nature, meaning they will store a unique copy of data for a longer period of time, maybe minutes or even hours. This is a very interesting development in storage and as we discuss in our article, Architecting Storage Networks for Data Delivery vs. Data Services, has the potential to relegate the current name brand NAS players to deliverers of software components further commoditizing mechanical storage systems.

Probably the more obvious payoff of using SSD as a large cache is helping to fix a performance problem more cost effectively than adding more mechanical drives or replacing those drives with faster drives. Less obvious is using these solutions on an initial purchase, they could very easily allow you to buy a mid-range SAN or NAS instead of a more expensive high end system.

Track us on Twitter: http://twitter.com/storageswiss

Subscribe to our RSS feed.

George Crump is lead analyst of Storage Switzerland, an IT analyst firm focused on the storage and virtualization segments. Find Storage Switzerland's disclosure statement here.

Comment  | 
Print  | 
More Insights
Register for Dark Reading Newsletters
White Papers
Cartoon
Current Issue
Dark Reading, September 16, 2014
Malicious software is morphing to be more targeted, stealthy, and destructive. Are you prepared to stop it?
Flash Poll
Title Partner’s Role in Perimeter Security
Title Partner’s Role in Perimeter Security
Considering how prevalent third-party attacks are, we need to ask hard questions about how partners and suppliers are safeguarding systems and data.
Video
Slideshows
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2012-5700
Published: 2014-09-22
Multiple cross-site scripting (XSS) vulnerabilities in Baby Gekko before 1.2.2f allow remote attackers to inject arbitrary web script or HTML via the (1) id parameter to admin/index.php or the (2) username or (3) password parameter in blocks/loginbox/loginbox.template.php to index.php. NOTE: some o...

CVE-2014-0484
Published: 2014-09-22
The Debian acpi-support package before 0.140-5+deb7u3 allows local users to gain privileges via vectors related to the "user's environment."

CVE-2014-2942
Published: 2014-09-22
Cobham Aviator 700D and 700E satellite terminals use an improper algorithm for PIN codes, which makes it easier for attackers to obtain a privileged terminal session by calculating the superuser code, and then leveraging physical access or terminal access to enter this code.

CVE-2014-3595
Published: 2014-09-22
Cross-site scripting (XSS) vulnerability in spacewalk-java 1.2.39, 1.7.54, and 2.0.2 in Spacewalk and Red Hat Network (RHN) Satellite 5.4 through 5.6 allows remote attackers to inject arbitrary web script or HTML via a crafted request that is not properly handled when logging.

CVE-2014-3635
Published: 2014-09-22
Off-by-one error in D-Bus 1.3.0 through 1.6.x before 1.6.24 and 1.8.x before 1.8.8, when running on a 64-bit system and the max_message_unix_fds limit is set to an odd number, allows remote attackers to cause a denial of service (dbus-daemon crash) or possibly execute arbitrary code by sending one m...

Best of the Web
Dark Reading Radio