This week, I will be in Auckland, New Zealand for the [IBM System x and System Storage Technical Symposium]. This is a three-day event, with 35 unique sessions and labs. The agenda is organized with a keynote session in the beginning, followed by 12 time slots over three days, each slot offering five different break-out session topics to choose from. Here is a recap of Day 1:
- Keynote Session
The keynote was led by Phil Tasker, IBM Business Unit Executive (BUE) for STG Education Programs in Growth Markets, then Matt Paterson, General Manager for Sales in New Zealand say a few words. IBM is in the Top 10 Training Hall of Fame, and conducts over 40,000 classes worldwide, resulting in over 1.3 million student days of instructions. IBM Systems Lab and Training technical hosts over three dozen conferences like this one every year. This is the first time that System x and Storage Symposium has been run in New Zealand, and based on the incredibly good turn-out, will probably be a regular event.
- Matt Ziegler - HPC
Matt Ziegler, IBM Senior HPC Solutions Architect for the iDataPlex marketing team, gave an introdcution to HPC during the keynote, then provided more details in a break-out session.
In the High Performance Computing (HPC) market, IBM POWER used to be the dominant chipset, with over 200 of the top 500 supercomputers back in June 2001. Today, only about 50 use POWER. Rather, over 350 of the top 500 supercomputers use x86 instead. HPC represents a 6.3 percent growth opportunity for computer, 9.3 percent growth for storage, and 8.6 percent growth for services.
IBM's leadership in energy efficiency applies to HPC as well. In the "Green 500", a ranking based on MFLOPS/Watt, 19 of the top 25 are from IBM. IBM's iDataPlex is the most energy efficient x86 platform, at 401 MFLOPS per Watt.
Overall, x86 is growing. In 2005, x86 had 48 percent of the market, RISC/Itanium had 39 percent, and mainframe had 12 percent. In 2009, x86 grew to 56 percent, RISC/Itanium dropped to 33 percent, and mainframe to 11 percent. By 2014, Matt projects that x86 will be 63 percent, RISC/Itanium will drop to 30 percent, and mainframe to 7 percent.
The most popular form factor for x86 are blades. Growing from 8 percent in 2005, to 20 percent in 2009, and projected to be 33 percent by 2014.
Recently, [IBM announced plans to acquire Platform Computing], a software company that makes a job workload scheduler for HPC environments. HPC and Cloud deployments are quite similar, and will share many technologies. Other newsworthy acquisitions include [Dell's acquisition of Force10 Networks] for Ethernet, and [Mellanox's acquisition of Voltaire] for Infiniband.
- IBM's Storage Strategy in the Era of Smarter Computing
I gave this presentation twice today. It has evolved quite a bit from the version I presented in Orlando last July. Attendees appreciated that my colorful analogies and stories helped them better understand the concepts of Big Data analytics, Workload-Optimized systems, and Cloud Storage offerings.
- SONAS Product Review and Demo
Rich Swain presented IBM's Scale-Out Network Attached Storage (SONAS) and provided a live demo connecting to a box here in New Zealand. This is a topic I often present at the Tucson Executive Briefing Center, but it is always good to hear someone else's spin.
- Welcome Reception
Phil Tasker invited everyone to the Welcome Reception after the last sessions. There was food and drink, and prizes! One person won an Xbox-360 game console, and two people won iPads.
technorati tags: IBM, Phil Tasker, Matt Paterson, Matt Ziegler, Rich Swain, HPC, Cloud, SONAS, Storage Strategy, Smarter Computing
Well, it's Tuesday, in the United States at least, and you know what that means... IBM Announcements! I am actually down under in Sydney, Australia, and it is Wednesday already as I write this. I feel like a time traveler.
IBM announces their latest disk system, the [IBM System Storage DCS3700], designed for high-performance computing (HPC), business analytics, video broadcasting, and other sequential workloads. The "DCS" stands for Deep Computing Storage. IBM already has the DCS9900 for large enterprise deployments, so this smaller DCS3700 is targeted for midrange deployments.
In a compact 4U package, the DCS3700 packs dual active-active controllers and up to 60 disk drives. The controller drawer can support two additional expansion drawers, of 60 drives each in 4U drawers, for a maximum total of 180 drives in 12U of rack space. Packed with "green" 7200RPM energy-efficient 2TB drives, a system can have up to a 360TB raw capacity. The system supports RAID levels 0, 1, 3, 5, 6, and 10.
The system comes with the latest 6Gbps SAS connections for host attachment, but you can choose 8Gbps Fibre Channel Protocol (FCP) instead, allowing the DCS3700 to be managed by SVC or Storwize V7000.
technorati tags: IBM, DCS3700, HPC, DCS, FCP
Wrapping up my week's theme of storage optimization, I thought I would help clarify the confusion between data reduction and storage efficiency. I have seen many articles and blog posts that either use these two terms interchangeably, as if they were synonyms for each other, or as if one is merely a subset of the other.
- Data Reduction is LOSSY
By "Lossy", I mean that reducing data is an irreversible process. Details are lost, but insight is gained. In his paper, [Data Reduction Techniques", Rajana Agarwal defines this simply:
"Data reduction techniques are applied where the goal is to aggregate or amalgamate the information contained in large data sets into manageable (smaller) information nuggets."
Data reduction has been around since the 18th century.
Take for example this histogram from [SearchSoftwareQuality.com]. We have reduced ninety individual student scores, and reduced them down to just five numbers, the counts in each range. This can provide for easier comprehension and comparison with other distributions.
The process is lossy. I cannot determine or re-create an individual student's score from these five histogram values.
This next example, complements of [Michael Hardy], represents another form of data reduction known as ["linear regression analysis"]. The idea is to take a large set of data points between two variables, the x axis along the horizontal and the y axis along the vertical, and find the best line that fits. Thus the data is reduced from many points to just two, slope(a) and intercept(b), resulting in an equation of y=ax+b.
The process is lossy. I cannot determine or re-create any original data point from this slope and intercept equation.
In this last example, from [Yahoo Finance], reduces millions of stock trades to a single point per day, typically closing price, to show the overall growth trend over the course of the past year.
The process is lossy. Even if I knew the low, high and closing price of a particular stock on a particular day, I would not be able to determine or re-create the actual price paid for individual trades that occurred.
- Storage Efficiency is LOSSLESS
By contrast, there are many IT methods that can be used to store data in ways that are more efficient, without losing any of the fine detail. Here are some examples:
- Thin Provisioning: Instead of storing 30GB of data on 100GB of disk capacity, you store it on 30GB of capacity. All of the data is still there, just none of the wasteful empty space.
- Space-efficient Copy: Instead of copying every block of data from source to destination, you copy over only those blocks that have changed since the copy began. The blocks not copied are still available on the source volume, so there is no need to duplicate this data.
- Archiving and Space Management: Data can be moved out of production databases and stored elsewhere on disk or tape. Enough XML metadata is carried along so that there is no loss in the fine detail of what each row and column represent.
- Data Deduplication: The idea is simple. Find large chunks of data that contain the same exact information as an existing chunk already stored, and merely set a pointer to avoid storing the duplicate copy. This can be done in-line as data is written, or as a post-process task when things are otherwise slow and idle.
When data deduplication first came out, some lawyers were concerned that this was a "lossy" approach, that somehow documents were coming back without some of their original contents. How else can you explain storing 25PB of data on only 1PB of disk?
(In some countries, companies must retain data in their original file formats, as there is concern that converting business documents to PDF or HTML would lose some critical "metadata" information such as modificatoin dates, authorship information, underlying formulae, and so on.)
Well, the concern applies only to those data deduplication methods that calculate a hash code or fingerprint, such as EMC Centera or EMC Data Domain. If the hash code of new incoming data matches the hash code of existing data, then the new data is discarded and assumed to be identical. This is rare, and I have only read of a few occurrences of unique data being discarded in the past five years. To ensure full integrity, IBM ProtecTIER data deduplication solution and IBM N series disk systems chose instead to do full byte-for-byte comparisons.
- Compression: There are both lossy and lossless compression techniques. The lossless Lempel-Ziv algorithm is the basis for LTO-DC algorithm used in IBM's Linear Tape Open [LTO] tape drives, the Streaming Lossless Data Compression (SLDC) algorithm used in IBM's [Enterprise-class TS1130] tape drives, and the Adaptive Lossless Data Compression (ALDC) used by the IBM Information Archive for its disk pool collections.
Last month, IBM announced that it was [acquiring Storwize. It's Random Access Compression Engine (RACE) is also a lossless compression algorithm based on Lempel-Ziv. As servers write files, Storwize compresses those files and passes them on to the destination NAS device. When files are read back, Storwize retrieves and decompresses the data back to its original form.
To read independent views on IBM's acquisition, read Lauren Whitehouse (ESG) post [Remote Another Chair, Chris Mellor (The Register) article [Storwize Swallowed], or Dave Raffo (SearchStorage.com) article [IBM buys primary data compression].
As with tape, the savings from compression can vary, typically from 20 to 80 percent. In other words, 10TB of primary data could take up from 2TB to 8TB of physical space. To estimate what savings you might achieve for your mix of data types, try out the free [Storwize Predictive Modeling Tool].
So why am I making a distinction on terminology here?
Data reduction is already a well-known concept among specific industries, like High-Performance Computing (HPC) and Business Analytics. IBM has the largest marketshare in supercomputers that do data reduction for all kinds of use cases, for scientific research, weather prediction, financial projections, and decision support systems. IBM has also recently acquired a lot of companies related to Business Analytics, such as Cognos, SPSS, CoreMetrics and Unica Corp. These use data reduction on large amounts of business and marketing data to help drive new sources of revenues, provide insight for new products and services, create more focused advertising campaigns, and help understand the marketplace better.
There are certainly enough methods of reducing the quantity of storage capacity consumed, like thin provisioning, data deduplication and compression, to warrant an "umbrella term" that refers to all of them generically. I would prefer we do not "overload" the existing phrase "data reduction" but rather come up with a new phrase, such as "storage efficiency" or "capacity optimization" to refer to this category of features.
IBM is certainly quite involved in both data reduction as well as storage efficiency. If any of my readers can suggest a better phrase, please comment below.
technorati tags: IBM, data reduction, storage efficiency, histogram, linear regression, thin provisioning, data deduplication, lossy, lossless, EMC, Centera, hash collisions, Information Archive, LTO, LTO-DC, SLDC, ALDC, compression, deduplication, Storwize, supercomputers, HPC, analytics
Continuing on the [IBM Storage Launch of February 9], John Sing has offered to write the following guest post about the [announcement] of IBM Scale Out Network Attached Storage [IBM SONAS]. John and I have known each other for a while, traveled the world to work with clients and speak at conferences. He is an Executive IT Consultant on the SONAS team.
[John Sing profile]
Guest Post written by John Sing, IBM San Jose, California
What is IBM SONAS? It’s many things, so let’s start with this list:
- It’s IBM’s delivery of a productized, pre-packaged Scale Out NAS global virtual file server, delivered in a easy-to-use appliance
- IBM’s solution for large enterprise file-based storage requirements, where massive scale in capacity and extreme performance is required, especially for today’s modern analytics-based Competitive Advantage IT applications
- Scales to many petabytes of usable storage and billions of files in a single global namespace
- Provides integrated central management, central deployment of petabyte levels of storage
- Modular commercial-off-the-shelf [COTS] building blocks. I/O, storage, network capacity scale independently of each other. Up to 30 interface nodes and 60 storage nodes, in an IBM General Parallel File System [GPFS]-based cluster. Each 10Gb CEE interface node port is capable of streaming at 900 MB/sec
- Files are written in block-sized chunks, striped over as many multiple disk drives in parallel – aggregating throughput on a massive scale (both read and write), as well as providing auto-tuning, auto-balancing
- Functionality delivered via one program product, IBM SONAS Software, which provides all of above functions, along with clustered CIFS, NFS v2/v3 with session auto-failover, FTP, high availability, and more
IBM SONAS makes automated tiered storage achievable and realistic at petabyte levels:
- Integrated global policy engine file placement, physical management, replication, deletion, archival
- Integrated high performance parallel scan engine capable of identifying files at over 10 million files per minute per node
- Integrated parallel data movement engine to physically relocate the data within tiered storage
And we’re just scratching the surface. IBM has plans to deploy additional protocols, storage hardware options, and software features.
However, the real question of interest should be, “who really needs that much storage capacity and throughput horsepower?”
The answer may surprise you. IMHO, the answer is: almost any modern enterprise that intends to stay competitive. Hmmm…… Consider this: the reason that IT exists today is no longer to simply save cost (that may have been true 10 years ago). Everyone is reducing cost… but how much competitive advantage is purchased through “let’s cut our IT budget by 10% this year”?
Notice that in today’s world, there are (many) bright people out there, changing our world every day through New Intelligence Competitive Advantage analytics-based IT applications such as real time GPS traffic data, real time energy monitoring and redirection, real time video feed with analytics, text analytics, entity analytics, real time stream computing, image recognition applications, HDTV video on demand, etc. Think of how GPS industry, cell phone / Twitter / Facebook, iPhone and iPad applications, as examples, are creating whole new industries and markets almost overnight.
Then start asking yourself, “What's behind these Competitive Advantage IT applications – as they are the ones that are driving all my storage growth? Why do they need so much storage? What do those applications mean for my storage requirements?”
( Read more about Analytics, pages 14-48 of [New Intelligence
for a Smarter Planet - Driving Business Innovation with IBM Analytic Solutions] )
To be “real-time”, long-held IT paradigms are being broken every day. Things like “data proximity”: we can no longer can extract terabytes of data from production databases and load them to a data warehouse – where’s the “real-time” in that? Instead, today’s modern analytics-based applications demand:
- Multiple processes and servers (sometimes numbering in the 100s) simultaneously ….
- Running against hundreds of terabytes of data of live production data, streaming in from expanding number of smarter sensors, input devices, users
- Producing digital image-intensive results that must be programatically sent to an ever increasing number of mobile devices in geographically dispersed storage
- Requiring parallel performance levels, that used to be the domain only of High Performance Computing (HPC)
This is a major paradigm shift in storage – and that is the solution and storage capabilities that IBM SONAS is designed to address. And of course, you should be able to save significant cost through the SONAS global virtual file server consolidation and virtualization as well.
Certainly, this topic warrants more discussion. If you found it interesting, contact me, your local IBM Business Partner or IBM Storage rep to discuss Competitive Advantage IT applications and SONAS further.
technorati tags: , IBM, Scale-Out, NAS, SONAS, John Sing, San+Jose, COTS, Petabytes, CEE, CIFS, NFS, FTP, HPC, New Intelligence, Analytics, Competitive Advantage IT Applications