Click here to close now.

Welcome!

Open Source Cloud Authors: Ruxit Blog, Liz McMillan, Elizabeth White, Lacey Thoms, Pat Romanski

Blog Feed Post

Don’t Buy RAID cards for SSD Caching

“Where have you been Dog?”

The IT Dog here again, back from summer sabbatical. Well, it was more like an “election” sabbatical since I spent the last six months working on the campaign to elect Barkington T. Howl, III as President of United Dogs of America (UDA). “Bark,” as his frat brothers call him, turned out to be a little too high-brow and out of touch with the rank-and-file UDA constituency. He lost in a landslide. Anyway, I am now back at my day job, talking about SSDs in the IT marketplace. Did you miss me?

Back to Our Regularly Scheduled Blog Topic

So what is the first thing to write about since coming back? How about looking at a trend in the RAID marketplace – offering RAID controller cards sold with proprietary SSD caching software designed to boost performance over traditional RAID offerings.

SSD Caching RAID Controller Cards

Redundant Array of Independent Disks or RAIDs, as they are affectionately known, have been around for many years. They are a vital component in many IT installations offering data redundancy and performance improvements over standard disk array. With the introduction of SSDs and SSD caching, many RAID controller card manufactures have updated their product offering to include the ability to run SSD caching algorithms on the RAID controller card itself. Examples of this in the market place include LSI MegaRAID controller card with CacheCade SSD caching software and Adaptec Series 7 controller cards with maxCache SSD caching software. The basic idea here is to buy the controller card from a particular vendor and use the SSD caching software they offer that runs only on their controller card.

Show Me The Money

Ok, so I understand the idea. Let’s see if we can figure out if this is the right way to do SSD caching. I am going to talk about LSI’s solution, not because “LSI” is easier to type than “Adaptec”, but because there happen to be some independent test results published on the web by Demartek for LSI MegaRAID with CacheCade. I am going to try and decipher just what the results are telling me. You can click on the link to read the entire report which documents test set-up etc. I am just going to discuss one chart presented to see what the fuss is about.

Figure 1 below is from page 8 of the report. It shows throughput in Megabits per second for a 90 minute web server test for the baseline system with no SSD caching and for the same system with SSD caching using one or two Intel X25-E 32GB SLC SSDs. The chart shows the baseline system without SSD caching maxed out at about 58 Mbps and that using one SSD and SSD caching, the performance improved to approximately 211 Mbps. Pretty nice. 3.6X improvement. And, with 2 SSDs for caching the throughput improved to 416 Mbps. 7.1X improvement! Excellent.

But let me dig into this a little more. The first thing I am trying to understand is just how they could get only 58 Mbps out of the baseline system. Remember now this is MegaBITS per second, not MegaBYTES per second. I am not a RAID controller guru, but I would have expected the baseline performance to be greater than 7.25 Mbytes/s. A quick internet search of RAID performance led me to this ZDNet page which listed a test RAID performance in MegaBYTES/s ranging from 64 to 257. So the baseline figure for the LSI test is suspect and therefore 3.6X or 7.1X improvement of really bad performance is not that impressive. But, if we ignore ___X improvement and just look at the data – 211 Mbits/sec (that’s 26 Mbytes/sec to you and me) is nothing to be too excited about.

Figure 1. LSI MegaRAID with CaceCade Demartek Throughput. Source: http://www.demartek.com/Reports_Free/Demartek_LSI_CacheCade_Performance_...

Velobit SSD Caching Software Results

So, I was trying to come up with an apples-to-apples comparison for the RAID SSD caching test data shown above but unfortunately I don’t have data which exactly duplicates the LSI test. However, Demartek performed system testing using Velobit HyperCache SSD caching software under different conditions which can be used to provide some general comparison observations.

Figure 2. below is taken from page 12 the Velobit Demartek report. The chart title is confusing: “Average MBPS – Linux without RD2”. When translated to English, the title means: Average MegaBYTES/s of a Linux based system running the vdbench workload test. The ‘without RD2’ part of the title means that there is a second chart in the report with a test called ‘RD2’ (mostly read operations test) whose results (more than 3200 Mbytes/s) cause the other test results shown in the first graph on P12 to be compressed and difficult to read. So the RD2 test results are removed for this graph (see the report) to zoom in on RD1, RD3 and RD4 results. Well, that was a lot to explain for a chart title, sorry about that.

Anyway, several takeaways (don’t you just hate that word) from this data:
1. The baseline performance for each test ranged from approx. 20-35 Mbytes/s (Mbytes!). Much higher than the baseline for the LSI tests (7.25 Mbytes/s).
2. The Velobit Hypercache performance ranged from 200 to 275 Mbytes/s (that is 1600 - 2200 Mbits/s to keep the same units as the LSI results)
3. These tests results also show testing results of FlashSoft SSD caching software because FlashSoft was available at Demartek for comparison purposes. By the way, FlashSoft outperforms the LSI CacheCade software significantly also.


Figure 2. Velobit HyperCache SSD Caching Software Demartek Test Results. Source: http://www.velobit.com/Portals/106427/docs/demartek_velobit_ssd_caching_...

Conclusion

This was a long and winding road to try to make a simple point. RAID systems and SSD caching are two independent components within your IT system. If you have to go with a RAID solution for redundancy reasons, don’t get lured into the illusion that you can solve two problems with one product: you do need to have your RAID solution also be your SSD caching solution. It may seem easy to combine RAID and SSD caching for you, but if you try to solve two independent ‘problems’ with one product, you may not be getting the best performance for either problem. SSD caching enabled RAID controller cards:

  • tend to be more expensive than standard RAID controller cards
  • limit your SSD caching software solutions to software compatible with that card (vendor lock-in)
  • do not perform as well as other SSD caching options


The benefit of using RAID based SSD caching is the software runs on the RAID card and does not consume any server CPU/memory resources. However, the benefits don’t seem to be worth the cost and performance hit you take by using this solution.

Read the original blog entry...

More Stories By Peter Velikin

Peter Velikin has 12 years of experience creating new markets and commercializing products in multiple high tech industries. Prior to VeloBit, he was VP Marketing at Zmags, a SaaS-based digital content platform for e-commerce and mobile devices, where he managed all aspects of marketing, product management, and business development. Prior to that, Peter was Director of Product and Market Strategy at PTC, responsible for PTC’s publishing, content management, and services solutions. Prior to PTC, Peter was at EMC Corporation, where he held roles in product management, business development, and engineering program management.

Peter has an MS in Electrical Engineering from Boston University and an MBA from Harvard Business School.

@ThingsExpo Stories
SYS-CON Events announced today that BMC will exhibit at SYS-CON's 16th International Cloud Expo®, which will take place on June 9-11, 2015, at the Javits Center in New York City, NY. BMC delivers software solutions that help IT transform digital enterprises for the ultimate competitive business advantage. BMC has worked with thousands of leading companies to create and deliver powerful IT management services. From mainframe to cloud to mobile, BMC pairs high-speed digital innovation with robust IT industrialization – allowing customers to provide amazing user experiences with optimized IT per...
As the Internet of Things unfolds, mobile and wearable devices are blurring the line between physical and digital, integrating ever more closely with our interests, our routines, our daily lives. Contextual computing and smart, sensor-equipped spaces bring the potential to walk through a world that recognizes us and responds accordingly. We become continuous transmitters and receivers of data. In his session at @ThingsExpo, Andrew Bolwell, Director of Innovation for HP's Printing and Personal Systems Group, discussed how key attributes of mobile technology – touch input, sensors, social, and ...
All major researchers estimate there will be tens of billions devices - computers, smartphones, tablets, and sensors - connected to the Internet by 2020. This number will continue to grow at a rapid pace for the next several decades. With major technology companies and startups seriously embracing IoT strategies, now is the perfect time to attend @ThingsExpo, June 9-11, 2015, at the Javits Center in New York City. Learn what is going on, contribute to the discussions, and ensure that your enterprise is as "IoT-Ready" as it can be
The Internet of Things is not only adding billions of sensors and billions of terabytes to the Internet. It is also forcing a fundamental change in the way we envision Information Technology. For the first time, more data is being created by devices at the edge of the Internet rather than from centralized systems. What does this mean for today's IT professional? In this Power Panel at @ThingsExpo, moderated by Conference Chair Roger Strukhoff, panelists will addresses this very serious issue of profound change in the industry.
WebRTC defines no default signaling protocol, causing fragmentation between WebRTC silos. SIP and XMPP provide possibilities, but come with considerable complexity and are not designed for use in a web environment. In his session at @ThingsExpo, Matthew Hodgson, technical co-founder of the Matrix.org, discussed how Matrix is a new non-profit Open Source Project that defines both a new HTTP-based standard for VoIP & IM signaling and provides reference implementations.
Buzzword alert: Microservices and IoT at a DevOps conference? What could possibly go wrong? In this Power Panel at DevOps Summit, moderated by Jason Bloomberg, the leading expert on architecting agility for the enterprise and president of Intellyx, panelists will peel away the buzz and discuss the important architectural principles behind implementing IoT solutions for the enterprise. As remote IoT devices and sensors become increasingly intelligent, they become part of our distributed cloud environment, and we must architect and code accordingly. At the very least, you'll have no problem fil...
"People are a lot more knowledgeable about APIs now. There are two types of people who work with APIs - IT people who want to use APIs for something internal and the product managers who want to do something outside APIs for people to connect to them," explained Roberto Medrano, Executive Vice President at SOA Software, in this SYS-CON.tv interview at Cloud Expo, held Nov 4–6, 2014, at the Santa Clara Convention Center in Santa Clara, CA.
Almost everyone sees the potential of Internet of Things but how can businesses truly unlock that potential. The key will be in the ability to discover business insight in the midst of an ocean of Big Data generated from billions of embedded devices via Systems of Discover. Businesses will also need to ensure that they can sustain that insight by leveraging the cloud for global reach, scale and elasticity.
In their session at @ThingsExpo, Shyam Varan Nath, Principal Architect at GE, and Ibrahim Gokcen, who leads GE's advanced IoT analytics, focused on the Internet of Things / Industrial Internet and how to make it operational for business end-users. Learn about the challenges posed by machine and sensor data and how to marry it with enterprise data. They also discussed the tips and tricks to provide the Industrial Internet as an end-user consumable service using Big Data Analytics and Industrial Cloud.
Building low-cost wearable devices can enhance the quality of our lives. In his session at Internet of @ThingsExpo, Sai Yamanoor, Embedded Software Engineer at Altschool, provided an example of putting together a small keychain within a $50 budget that educates the user about the air quality in their surroundings. He also provided examples such as building a wearable device that provides transit or recreational information. He then reviewed the resources available to build wearable devices at home including open source hardware, the raw materials required and the options available to power s...
How do APIs and IoT relate? The answer is not as simple as merely adding an API on top of a dumb device, but rather about understanding the architectural patterns for implementing an IoT fabric. There are typically two or three trends: Exposing the device to a management framework Exposing that management framework to a business centric logic Exposing that business layer and data to end users. This last trend is the IoT stack, which involves a new shift in the separation of what stuff happens, where data lives and where the interface lies. For instance, it's a mix of architectural styles ...
We certainly live in interesting technological times. And no more interesting than the current competing IoT standards for connectivity. Various standards bodies, approaches, and ecosystems are vying for mindshare and positioning for a competitive edge. It is clear that when the dust settles, we will have new protocols, evolved protocols, that will change the way we interact with devices and infrastructure. We will also have evolved web protocols, like HTTP/2, that will be changing the very core of our infrastructures. At the same time, we have old approaches made new again like micro-services...
Connected devices and the Internet of Things are getting significant momentum in 2014. In his session at Internet of @ThingsExpo, Jim Hunter, Chief Scientist & Technology Evangelist at Greenwave Systems, examined three key elements that together will drive mass adoption of the IoT before the end of 2015. The first element is the recent advent of robust open source protocols (like AllJoyn and WebRTC) that facilitate M2M communication. The second is broad availability of flexible, cost-effective storage designed to handle the massive surge in back-end data in a world where timely analytics is e...
Collecting data in the field and configuring multitudes of unique devices is a time-consuming, labor-intensive process that can stretch IT resources. Horan & Bird [H&B], Australia’s fifth-largest Solar Panel Installer, wanted to automate sensor data collection and monitoring from its solar panels and integrate the data with its business and marketing systems. After data was collected and structured, two major areas needed to be addressed: improving developer workflows and extending access to a business application to multiple users (multi-tenancy). Docker, a container technology, was used to ...
The true value of the Internet of Things (IoT) lies not just in the data, but through the services that protect the data, perform the analysis and present findings in a usable way. With many IoT elements rooted in traditional IT components, Big Data and IoT isn’t just a play for enterprise. In fact, the IoT presents SMBs with the prospect of launching entirely new activities and exploring innovative areas. CompTIA research identifies several areas where IoT is expected to have the greatest impact.
The Industrial Internet revolution is now underway, enabled by connected machines and billions of devices that communicate and collaborate. The massive amounts of Big Data requiring real-time analysis is flooding legacy IT systems and giving way to cloud environments that can handle the unpredictable workloads. Yet many barriers remain until we can fully realize the opportunities and benefits from the convergence of machines and devices with Big Data and the cloud, including interoperability, data security and privacy.
The Internet of Things is tied together with a thin strand that is known as time. Coincidentally, at the core of nearly all data analytics is a timestamp. When working with time series data there are a few core principles that everyone should consider, especially across datasets where time is the common boundary. In his session at Internet of @ThingsExpo, Jim Scott, Director of Enterprise Strategy & Architecture at MapR Technologies, discussed single-value, geo-spatial, and log time series data. By focusing on enterprise applications and the data center, he will use OpenTSDB as an example t...
Scott Jenson leads a project called The Physical Web within the Chrome team at Google. Project members are working to take the scalability and openness of the web and use it to talk to the exponentially exploding range of smart devices. Nearly every company today working on the IoT comes up with the same basic solution: use my server and you'll be fine. But if we really believe there will be trillions of these devices, that just can't scale. We need a system that is open a scalable and by using the URL as a basic building block, we open this up and get the same resilience that the web enjoys.
We are reaching the end of the beginning with WebRTC, and real systems using this technology have begun to appear. One challenge that faces every WebRTC deployment (in some form or another) is identity management. For example, if you have an existing service – possibly built on a variety of different PaaS/SaaS offerings – and you want to add real-time communications you are faced with a challenge relating to user management, authentication, authorization, and validation. Service providers will want to use their existing identities, but these will have credentials already that are (hopefully) i...
Container frameworks, such as Docker, provide a variety of benefits, including density of deployment across infrastructure, convenience for application developers to push updates with low operational hand-holding, and a fairly well-defined deployment workflow that can be orchestrated. Container frameworks also enable a DevOps approach to application development by cleanly separating concerns between operations and development teams. But running multi-container, multi-server apps with containers is very hard. You have to learn five new and different technologies and best practices (libswarm, sy...