SlideShare ist ein Scribd-Unternehmen logo
1 von 14
Downloaden Sie, um offline zu lesen
TheBathwickGroup


      A Bathwick Group
              white paper




                                             The mainframe
                                             and the cloud

                The role of the
            mainframe as a
       host environment
                    for enterprise
          cloud computing


                                                Document Number:   BC-INF-W-00068UK-EN-00
All contents © The Bathwick Group Ltd 2010      Author:            Gary Barne!
                                                Published:         November 2010
TheBathwickGroup
        The mainframe and the cloud - November 2010




        Abstract
        Nearly all of the discussion and debate surrounding cloud compu!ng presumes that the ideal enterprise cloud
        environment will be built on x86 processors in a massively scaled-out architecture. Un!l recently there has
        been very li"le discussion about the poten!al benets that the mainframe o#ers the enterprise as a host for
        a cloud compu!ng environment.
        This paper presents the poten!al benets that enterprises could derive from the mainframe as a cloud
        environment in terms of cost, reliability, scalability, security and exibility. We also set out to highlight the
        changes that have taken place within the mainframe eco-system that mean many of the tradi!onal cri!cisms
        of the mainframe pla$orm are simply out of date.

        Key facts
        The latest version of the mainframe introduces:
        !   "     A consolida"on pla#orm that could reduce energy consump"on, oor space required, labour costs
                  and so$ware licensing costs by as much as 90%
        !   "     A virtualisa"on environment that can support 1000s of mainframe, Linux, and Java environments,
                  all within one physical machine
        !   "     An environment within which System p blades (running AIX) and – in the rst half of 2011 - Linux
                  running on System x blades can run directly alongside the mainframe using the same virtual
                  network and bene%ng from similar levels of reliability, availability and scalability as the mainframe
                  itself.

        Key messages
            !
            "     Contrary to popular myth, the mainframe is alive and well. We o$en hear commentators claiming
                  that the era of the mainframe is “over”, despite the fact that IBM’s mainframe franchise con"nues
                  to grow. In one sense, these statements are true; the mainframe of the 1980s is dead. The 80s
                  mainframe has been replaced by a pla#orm that is more open, more cost e&ec"ve and more
                  exible than ever before.
        !   "     The mainframe represents a compelling proposi!on as a cloud hos!ng environment. The
                  mainframe o&ers an alterna"ve to the massively scaled-out environments that are currently being
                  touted as the future of cloud. Scale-out architectures undoubtedly bring advantages, and in some
                  cases they certainly do represent the best approach, but the scale-up architecture of the mainframe
                  brings its own benets that mean that for some workloads it o&ers signicant advantages over
                  scaled-out systems. The key lies in understanding which situa"ons indicate one approach or
                  the other.
            "!    Dispelling the myths of complexity, integra!on and cost. Cri"cs of the mainframe typically cite the
                  same set of problems with the mainframe pla#orm – “It is expensive”, “It is a closed environment”,
                  “It is hard to develop so$ware for the mainframe”. These cri"cisms may have been true 15 years
                  ago but in the intervening "me the mainframe has changed radically.




All contents © The Bathwick Group Ltd 2010                      2                                               www.bathwick.com
TheBathwickGroup
        The mainframe and the cloud - November 2010




            "!    IBM’s latest mainframe transforms the value proposi!on. The mainframe value proposi"on has
                  been undergoing a con"nuous transforma"on over the past two decades in terms of the
                  so$ware stack and the hardware. The latest incarna"on of the mainframe, the zEnterprise, takes
                  this evolu"on a signicant way further by making it possible to combine mainframe and non-
                  mainframe hardware within a single managed environment.
            "!    If you’re considering consolida!ng SAP, your web infrastructure or your BI pla"orms, consider
                  the mainframe. IBM has focussed on exactly these applica"on scenarios in developing the
                  new mainframe and even organisa"ons that don’t currently have a mainframe might nd the
                  overall value proposi"on compelling.

        Dispelling some mainframe myths
        The next "me someone says that the mainframe is dead or that the era of the mainframe is over, ask them
        what compu"ng pla#orm manages their credit-card payments, controls the ATM network they use to withdraw
        cash, or underpins the globe’s airline "cke"ng systems; almost without excep"on it’s a mainframe.
        In the 1980s the mainframe was perceived as costly, complex, and closed. There was some jus"ca"on for
        this view and it helped to drive the adop"on of client-server compu"ng as a cheaper, less complex and more
        open approach to compu"ng. However, over the past two decades, the mainframe has undergone a drama"c
        transforma"on aimed at directly addressing the shortcomings of the 80s mainframe.

        The mainframe of the 1980s is dead; welcome to the mainframe of the 2010s
        In the 1980s client-server compu"ng emerged as a response to the cost and complexity associated with the
        mainframe. Client-server compu"ng promised lower cost, greater exibility, and an environment that was far
        more open in terms of the ease with which di&erent systems could be integrated with one another. When
        you look back at the mainframe of the 1980s it’s easy to see why client-server compu"ng was regarded as a
        be!er, less expensive alterna"ve.

        1980s – Closed, proprietary, costly


                                               Mainframe             Distributed




                          Figure 1.          The mainframe of the 1980s




All contents © The Bathwick Group Ltd 2010                                3                                www.bathwick.com
TheBathwickGroup
        The mainframe and the cloud - November 2010




        In the 1980s the mainframe was a fortress. On one hand, it was secure, reliable and scalable, but on the
        other hand it was expensive and inaccessible. Companies that wanted to integrate mainframe data with their
        emerging client-server applica"ons o$en had to resort to ETL solu"ons that dumped the mainframe data
        onto tape so that it could then be uploaded into the client-server system.
        Within the mainframe environment, you were free to develop your own so$ware – provided you used
        mainframe programming languages like Fortran, COBOL or PL/1. Meanwhile, within the world of distributed
        compu"ng there was an explosion of innova"on in the world of programming languages, development tools
        and middleware.

        1990s – Fortress mainframe gets a drawbridge


                                              Mainframe              Distributed




                          Figure 2.          The mainframe of the 1990s



        In the 1990s IBM took the rst serious steps to make it easier
        to integrate mainframe-based applica"ons and non-mainframe
        applica"ons by introducing connec"vity middleware and a number
        of third-par"es (including the likes of Tuxedo and So$ware AG)              “ I predict that the last main-
        developed alterna"ve middleware products that made it possible              frame will be unplugged on
        (albeit at considerable expense and e&ort) to link the mainframe            March 15, 1996. “
        and non-mainframe worlds.                                                   Stewart Alsop, Infoworld March
                                                                                    1991
        This was undoubtedly “progress”, but the mainframe was s"ll an
        ocean away from being a genuinely rst class ci"zen in the world
        of distributed compu"ng. Developers s"ll had to use “mainframe”
        development tools and programming languages. “Mainframe” middleware and the mainframe was s"ll a
        complex and costly environment to manage and run. The apparent failure of the mainframe to respond to the
        challenges presented by client-server compu"ng coupled with the sheer volume of innova"on and crea"vity
        within the client-server world resulted in a general belief that the mainframe was a dying pla#orm. This belief
        prompted one commentator to declare in 1991 that “The last mainframe will be unplugged on March 15,
        1996”.




All contents © The Bathwick Group Ltd 2010                                4                                   www.bathwick.com
TheBathwickGroup
        The mainframe and the cloud - November 2010




        2000s – Reinven!ng the mainframe

                                               Distributed
                                               Mainframe




                          Figure 3.          The mainframe of the 2000s



        By the end of the 1990s, it was clear that IBM needed to take ac"on to re-invigorate the mainframe. In 1999
        IBM announced plans to work with RedHat to bring Linux to the mainframe pla#orm and was able to o&er
        mainframe Linux as a standard op"on in 2000.

        A    Y
        The 2000s saw a number of profound changes to the mainframe environment. The most important theme
        surrounding these changes was the decision to take a systems oriented approach to the ongoing development
        of the mainframe by integra"ng the e&orts of IBM’s so$ware and hardware groups. In 2003 IBM made a very
        public commitment to the mainframe by introducing the Mainframe Charter – a promise to its customers
        that IBM would con"nue to invest in the pla#orm by inves"ng in mainframe hardware design, the so$ware
        ecosystem and in making it easier to manage and integrate the pla#orm within a distributed compu"ng
        environment.
        Throughout the 2000s, major mainframe hardware announcements went hand in hand with so$ware
        announcements and the por#olio of mainframe so$ware products became increasingly aligned with IBM’s
        non-mainframe so$ware por#olio.
        The improvement this strategy brought about in terms of integra"ng IBM’s di&erent mainframe o&erings and
        integra"on between IBM’s non-mainframe products as well was signicant for two key reasons; it reduced
        the cost and complexity associated with provisioning a mainframe and helped IBM to o&er a more integrated
        alterna"ve to the host of third-party products that are available for the mainframe from vendors like CA and
        Compuware.




All contents © The Bathwick Group Ltd 2010                                5                                www.bathwick.com
TheBathwickGroup
        The mainframe and the cloud - November 2010




                  So#ware and hardware alignment

                  Throughout the 2000s IBM set about aligning its distributed and mainframe so$ware o&erings;

                  !
                  "!         CICS was enhanced to support web services with the introduc"on of SOAP for CICS

                  !
                  "!         IMS added support for RSS

                  !
                  "!         IBM’s mainframe development tools por#olio was integrated within the Ra"onal
                             development tools family (Ra"onal Developer for System z)

                  !
                  "!         IBM’s systems management so$ware Tivoli was integrated so that key components (like
                             iden"ty management, for example) could run on either distributed or mainframe systems

                  !
                  "!         DB2 was enhanced to improve the level of integra"on between the mainframe and non-
                             mainframe versions of the database

                  !
                  "!         Mainframe management was drama"cally simplied with the release of Systems Director
                             (a graphical management tool for the mainframe that automates a host of administra"on
                             tasks)




        The hardware con"nued to evolve with improvements to the price/performance of the machine itself. This
        was further enhanced with the introduc"on of the ability to o'oad specic workloads (notably Java and
        Linux) from the main processors onto less specialist processors that were less costly themselves and also
        reduced the number of “core” MIPS that had to be licensed. At the same "me key IBM so$ware products
        – DB2, Ra"onal development tools, the Websphere middleware family and the Tivoli systems management
        suite all became increasingly aligned with both the mainframe and distributed pla#orms.

        SOA    
        One of the key changes in the mainframe was the addi"on of much improved support for SOA. Venerable
        technologies like CICS and IMS were all extended to make it possible to integrate them within an SOA-based
        environment. CICS now ships with support for SOAP and web services. Technologies like MQ Series act as a
        SOA bridge between the mainframe and other environments and the combina"on of middleware, security
        technology and management tools mean that for some organisa"ons the mainframe can serve as a single
        organisa"on-wide ESB.




All contents © The Bathwick Group Ltd 2010                          6                                                www.bathwick.com
TheBathwickGroup
        The mainframe and the cloud - November 2010




        2010 – The next genera!on

                                                Distributed
                                                Mainframe




                          Figure 4.          The next genera!on of the IBM mainframe



        In July 2010, IBM launched the zEnterprise server, which delivered two sets of benets. Firstly, the new
        mainframe delivers signicant improvements to the basic performance of the mainframe:
            "!    The fastest general purpose CPU ever, running at 5.2Ghz
            "!    Up to 40% performance improvement for tradi"onal mainframe workloads
            "!    80% improvement in Linux price/performance
            "!    A consolida"on pla#orm that could reduce energy consump"on, oor space required, labour
                  costs and so$ware licensing costs by as much as 90%
            "!    Introduc"on of RAIM (redundant array of independent memory) which signicantly improves
                  the reliability of system memory
        Secondly, and perhaps the most signicant feature of this release, is the integra"on the pla#orm o&ers with
        non-mainframe hardware. The zEnterprise can be a!ached to a zBX, which provides housing for up to 112
        p-Series blades, running AIX; x-Series blades (running Linux) will be supported in the rst half of 2011. The zBX
        is connected to the zEnterprise machine via two private network connec"ons: a 10Gb Ethernet connec"on is
        used by the applica"ons to provide ultra highspeed connec"vity, and a second management network (running
        at 1Gb/s) is used by the machine’s management so$ware to control the di&erent hardware devices within
        the system.
        The blades housed within the zBX are managed, along with the mainframe itself, by a single management
        component called zManager. The zManager adds features to blade technology management that have
        tradi"onally been the sole preserve of the mainframe. The management component automa"cally discovers
        new hardware, manages and monitors running hardware, and supports running upgrades to so$ware and
        rmware.
        The zBX is built with levels of hardware redundancy comparable to the mainframe itself, so key components
        like power supplies, power distribu"on units, network switches and fans are all shipped with at least one
        addi"onal or redundant component.




All contents © The Bathwick Group Ltd 2010                                7                                     www.bathwick.com
TheBathwickGroup
        The mainframe and the cloud - November 2010




        Cloud in a box : Using the mainframe as the host for a cloud environment
        Cloud compu"ng is being heralded as the next evolu"on of systems architecture. The benets of cloud
        compu"ng are primarily built on the no"on that cloud environments allow you to gain benets in terms of
        cost and exibility by virtualizing the di&erent workloads and deploying them within an environment that can
        automa"cally scale-up or down as workload demands change over "me.
        Broadly speaking, cloud compu"ng environments can be;
            "!    Public
                   The environment is run on shared infrastructure accessible via the internet. In this scenario you pay
                   a third party like Amazon (EC2) to run your workload and are charged according to the CPU "me your
                   applica"ons use, the bandwidth they consume and the storage you require.
            "!    Semi-Private
                   The environment is run by a third party in much the same way as a tradi"onal outsourcing contract,
                   but you are charged on the basis of u"lisa"on rather than paying a at fee.
            "!    Private
                   The environment is run on your own premises. In this scenario you maintain and run your own cloud
                   infrastructure.

        C   
        While there are undoubtedly cases where public cloud is well suited to enterprise compu"ng, these cases
        tend to be where the required func"onality is well dened and has clear boundaries. IBM and others already
        o&er cloud-based development environments, for example, which allow development teams to quickly ramp
        up (or down) as their needs change.
        Public cloud is less suited to complex applica"ons where some of the processing or data can’t be run in the
        cloud. Imagine you have an ERP system running on your own infrastructure, and you want to deploy a web
        applica"on that is available to your customers. It is likely that you won’t be able to simply deploy your ERP
        applica"on in the cloud, so that will have to remain on your infrastructure.
        If you deploy the “web components” of the solu"on in a public cloud environment, you will have to manage
        the link between the cloud-based applica"on and your ERP system. While it is possible to do this, and you can
        do this securely by crea"ng a VPN that reaches beyond your infrastructure into the cloud environment, there
        is a real likelihood that you’ll encounter signicant performance headaches as transac"ons have to cross from
        the cloud environment, over the network, and into your infrastructure.
        If you could derive the same benets of scalability, ease of management, and automa"on by deploying cloud
        technology on your own infrastructure, you can avoid the problems that network latency creates – or at
        least you’ll be in a much be!er posi"on to manage them, as the whole applica"on will be running on your
        network.




All contents © The Bathwick Group Ltd 2010                      8                                              www.bathwick.com
TheBathwickGroup
        The mainframe and the cloud - November 2010




        Building your own cloud
        Many pundits will say that the only way to build a private cloud is to deploy a virtualisa"on technology, like
        VMware, and install a number of blades that will provide the processing horsepower for your cloud. But there
        is an alterna"ve to the conven"onal belief that “scale-out” is the way to deliver cloud compu"ng.



                                                           Scale-up            Scale-out




                          Figure 5.          Scale-up and Scale-out compu!ng



        Figure 5 shows, in very simple terms, the architectural di&erence between mainframe or scale-up compu"ng
        (where many images run within a single mul"-processor machine) and scale-out where one or more images run
        in a number of machines, that are then managed as a single en"ty by a virtualisa"on solu"on like VMware.
        Looking at the diagram above, it is worth asking the following ques"ons about the two approaches;
            "!    What is the likely di&erence in physical footprint?
            "!    What is the likely di&erence in power consump"on?
            "!    How quickly can you increase capacity?
            "!    What is the likely di&erence in reliability?
            "!    What is the likely di&erence in the "me it would take for a request to be sent from one virtual
                  image to another?
        Few proponents of scale-out compu"ng would claim that you can get more processing power per square foot
        of oor space by taking a scale-out approach; the total power consump"on (factoring in all power u"lised
        in order to operate the hardware – including cooling) per processor of a scale-out solu"on vs scale-up can
        be as much as 20 "mes more. Increasing capacity in a scale-out solu"on is o$en described as simply a case
        of “pushing another blade into a slot”, but that assumes that you have a blade that is wai"ng to be installed
        (which seems like a bit of a waste of capital and capacity) and that you have a vacant slot into which the new
        blade can be inserted. If you don’t have a blade standing by, or if you need to install a new blade enclosure




All contents © The Bathwick Group Ltd 2010                                9                                    www.bathwick.com
TheBathwickGroup
        The mainframe and the cloud - November 2010




        then adding capacity could take a considerable amount of "me. In the scale-up world, IBM can ship servers
        which are over-provisioned in terms of hardware, with the addi"onal processors ready to be enabled on-
        demand. Even if new hardware needs to be added, this can be requested as part of the standard procurement
        process and executed as part of the standard maintenance process.
        The ques"on of reliability is a complex one. Certainly, a catastrophic failure within an SMP based machine
        will result in a complete outage, while a catastrophic failure within a blade (or even a blade enclosure) may
        only result in a par"al outage for a subset of users. But it isn’t as simple as that. To begin with, the resilience
        of a scale-out applica"on to component failures depends, to a large degree, on the design of the applica"on.
        Some applica"ons have been designed to recover from a par"al failure while others may fail completely.
        In order to deliver high levels of reliability and availability in a scale-out environment, you need to specify
        mul"ple network devices, redundant power supplies, and middleware that supports fail-over and recovery. In
        the mainframe environment the same steps have been taken, but these are internalised within the machine
        itself. Taking a simple example like RAM (faulty RAM is one of the more common reasons for the failure of
        a blade) mainframe RAM is designed to be fault tolerant, faulty memory loca"ons can be quaran"ned and
        errors can be reported before they result in a system failure. With the latest release of the mainframe IBM
        has introduced RAIM – the RAM equivalent of RAID.
        The last ques"on depends on the laws of physics. If an electron has to travel 3 metres it will take 10 "mes longer
        to reach its des"na"on than one that has to travel 30 cen"metres. This ignores the di&erent layers of so$ware
        and network hardware that the electron (or request) may have to pass through on its journey. The mainframe
        supports virtual networking, which is signicantly faster than machine to machine communica"on.
        Both approaches carry their own benets and disadvantages – and the aim of this paper isn’t to say that
        scale-up is the only, and best, way to deliver internal cloud. But the idea that scale-out is the only, and best,
        way is equally untrue.

        A -  Y: MARSH
        Marsh is the world’s largest insurance broker and risk advisor,
        with 24,000 employees worldwide and annual revenues of
                                                                                 “ Our cost-benet analysis showed
        almost $5 billion. Its European, Middle Eastern and African
                                                                                 that a distributed architecture based
        (EMEA) business units have their headquarters in London,                 on x86 cores would be signicantly
        and the company employs approximately 11,000 people.                     more expensive than an IFL engine –
                                                                                 which shows that it’s denitely worth
        The company took the decision to modernize an exis"ng
                                                                                 inves!ga!ng the Enterprise Linux
        mainframe applica"on by developing a modern web-based
                                                                                 Server op!on even for rela!vely
        interface to it. The company chose to use IBM Ra"onal                    small environments. “
        Host Access Transforma"on Services (HATS) to support the                 Tim Benne! - Project Manager, Marsh
        web-enablement of the applica"on, and then set about
        determining the ideal hardware environment for the new
        func"onality.




All contents © The Bathwick Group Ltd 2010                      10                                                www.bathwick.com
TheBathwickGroup
        The mainframe and the cloud - November 2010




                                                                   The company had ini"ally intended to build a new
                                                                   Linux environment on standard x86-based servers,
         “ Compared to procuring several new Intel
         servers, wai!ng for them to be delivered,                 when IBM proposed using an Enterprise Linux Server
         nding space for them in the data centre,                 solu"on instead, using an Integrated Facility for Linux
         connec!ng all the cables and so on, it was a              processor in their System z9 Business Class server,
         really quick and easy process. We didn’t need             they conducted a comparison of both op"ons. The
         to buy any new hardware, and we completed                 conclusion of the study was that the mainframe-
         the whole implementa!on during a standard
                                                                   based solu"on o&ered be!er levels of reliability and
         weekend maintenance slot, so there was no
                                                                   exibility, at signicantly lower cost to the Intel based
         disrup!on for our users. ”
                                                                   proposal.
         Tim Benne! – Project Manager, Marsh
                                                            Since the company already had an IBM mainframe,
                                                            the deployment of the new infrastructure was simply
                                                            a ma!er of upgrading the machine, a task that was
        completed by an IBM engineer during a scheduled maintenance window.

        A -  Y:  UY  B
        The University of Bari (Universita di Bari) in Italy has chosen to build a cloud compu"ng service on top of an
        IBM System z mainframe running the Linux opera"ng system with DB2, WebSphere and Tivoli middleware
        from IBM.
        The goal of the project is to help local businesses, including shermen, winemakers and trucking companies,
        gain access to compu"ng power that would otherwise be unavailable to them.
        One example allows local shermen to determine demand in local sh markets in real "me, using a simple
        touch screen interface installed in their boats. The shermen enter the type of sh caught just minutes ago
        and instantaneously start a virtual auc"on with wholesalers on the docks. Another project links winemakers
        at up to 60 co-opera"ve wineries to enable them to determine market demand for various types of wines,
        then package and ship wines demanding the highest price.
        The overall project is being undertaken as part of the Daisy-Net consor"um, which combines public and
        private organisa"ons from ve regions of southern Italy. Daisy-Net researches, develops and provides transfer
        and training ac"vi"es for new technologies in the informa"on and communica"on technology sector and
        beyond.

        Introducing the Enterprise Linux Server – Cloud in a box
        The Enterprise Linux Server (ELS) is made up of an IBM z-Series mainframe that ships with the hardware and
        so$ware necessary to create and manage a private cloud of Linux servers. The system takes advantage of
        the mainframe’s built-in support for virtualisa"on, combined with IBM’s Tivoli Service Automa"on Manager,
        which provides the management and provisioning framework for the machine.
        The ELS can also be used to support Linux-based desktop virtualisa"on, using eyeOS, an open-source Virtual
        Linux Desktop pla#orm.




All contents © The Bathwick Group Ltd 2010                    11                                                   www.bathwick.com
TheBathwickGroup
        The mainframe and the cloud - November 2010




        The key features of the ELS are that it is built on mature
        technology, and makes extensive use of IBM’s so$ware
        por#olio (Tivoli, Ra"onal, Systems Director) to provide all of
        the core infrastructure that you need in order to deploy an            “ Allianz, a global insurance company
        internal cloud – all within a single hardware footprint.               based in Germany, e$ec!vely
                                                                               migrated applica!ons from 60 distrib-
        The system allows users to dene their own standard images
                                                                               uted systems to one Linux on IBM
        that can drama"cally reduce the "me it takes to deploy                 System z computer in 48 hours at its
        new servers. This approach can have a signicant impact on             Australian data centre. “
        hardware and so$ware costs (IBM has seen savings of up to
        75% in some scenarios) as well as reducing the cost and "me
        associated with systems management. Another by-product
        of using standardised images is the reduc"on in faults arising
        from congura"on errors (which in some environments can account for the vast majority of applica"on
        failures).

        T E L S    ATOS
        Atos Origin is a major interna"onal IT services provider, employing 50,000 professionals in 40 countries.
        The company o&ers integrated design, build and operate solu"ons to large mul"-na"onal clients in targeted
        industry sectors, and has annual revenues of around £5.8 billion. An important part of Atos Origin’s UK business
        is managed services – from basic server hos"ng through to full lifecycle management for both hardware and
        applica"ons.
                                                         Atos Origin recently developed a new Linux-based applica"on
                                                         that would help rail transport companies deliver travel
                                                         informa"on to sta& and passengers via their mobile phones,
         “ We have been using IFL processors             and wanted to deliver it as a shared service for several major
         for several years now, and in our
                                                         rail companies. The company also wanted to enhance its
         opinion, System z has proved itself as
                                                         service to customers in the insurance industry by upgrading
         an ideal enterprise Linux pla"orm. “
                                                         from MQSeries Integrator (MQSI) middleware to a new IBM
         Colin Clews - Technologies Manager, Atos
                                                         WebSphere Message Broker solu"on, which would provide
         Origin
                                                         support for Web Services integra"on for a number of core
                                                         insurance applica"ons.
                                                          Atos Origin decided to use the IBM Enterprise Linux Server
        (ELS) to host these systems. The ELS is an IBM System z10 Business Class server that leverages specialised IBM
        Integrated Facility for Linux (IFL) processors to provide a highly available environment for virtualised Linux
        workload.
        “We have been using IFL processors for several years now, and in our opinion, System z has proved itself as an
        ideal enterprise Linux pla#orm,” says Colin Clews. “With a tradi"onal distributed approach – even using the
        latest virtualisa"on-enabled x86 processors – you end up with a lot of unreliable boxes to manage.




All contents © The Bathwick Group Ltd 2010                     12                                              www.bathwick.com
TheBathwickGroup
        The mainframe and the cloud - November 2010




        With the ELS, we can run hundreds of environments
        within a single physical footprint, and easily deliver the       “ The z10 IFL engines in the new ELS not
        24x7 availability that our customers demand.”                    only run Linux environments approxi-
                                                                         mately twice as fast as the previous
        The addi"onal capacity of the z10 IFLs is available at a
                                                                         genera!on – they also o$er around 40
        50 percent lower price than the previous genera"on               percent more capacity, enabling us to
        of System z servers. The result is a price-performance           expand our Linux footprint without
        improvement of up to 65 percent.                                 increasing our costs. “
        Running the mobile applica"on under Linux on the ELS             Colin Clews - Technologies Manager, Atos

        proved so successful that Atos also decided to deploy            Origin

        the WebSphere Message Broker solu"on in a similar
        environment. The company has invested in a second IFL
        to handle this new workload.
        “Running WebSphere Message Broker under Linux on the ELS is considerably more cost-e&ec"ve than our
        exis"ng MQSI and z/OS environment,” says Colin Clews. “As a result, we not only benet from the extended
        func"onali"es that Message Broker o&ers – we can also deliver them at a more compe""ve price for our
        customers.”
        To provide disaster recovery for the System z environment, the produc"on ELS machine is mirrored to a
        smaller z9 server at a secondary data centre. The z9 server features IBM Capacity Backup Upgrade (CBU),
        which enables its capacity to be drama"cally increased if and when it is required to take over the produc"on
        workload from the produc"on server. With CBU, this spare capacity a!racts no maintenance or so$ware
        licensing costs un"l it is actually ac"vated.
        Linux on the mainframe is also helping Atos Origin develop an on-demand compu"ng model for its customers,
        enabling them to pay for compu"ng power on a transac"onal basis. “With z/VM, we can provision new Linux
        servers very rapidly, and measure capacity usage very precisely,” says Colin Clews. “If customers need more
        power, we can just turn it on almost instantly, and bill them accordingly.
        “We think that the on-demand exibility o&ered by Linux on the IBM ELS pla#orm will be very a!rac"ve
        to customers who currently run large distributed server environments, enabling them to reduce costs and
        increase responsiveness.”

        Summary
        It isn’t the inten"on of this paper to say that the mainframe is the “best” host for all cloud compu"ng
        environments, but that in some situa"ons we believe that it certainly is. Much of the current discourse around
        virtualisa"on and cloud compu"ng ignores the power of large scale-up environments like the mainframe
        and this is simply foolish. The mainframe of today is profoundly di&erent from the mainframe of the 1980s
        and when deba"ng the ideal pla#orm for cloud environments, it is only fair to compare today’s mainframe
        with non-mainframe technology rather than relying on a dated understanding of the mainframe’s value
        proposi"on.




All contents © The Bathwick Group Ltd 2010                     13                                            www.bathwick.com
TheBathwickGroup
        The mainframe and the cloud - November 2010


        About The Bathwick Group
        The Bathwick Group is a research-based consul"ng company that helps clients address their most pressing needs in strategic
        planning, go-to-market planning and execu"on, and IT infrastructure e&ec"veness:

        Consul!ng
            "     Bathwick Engage A rapid collabora"ve consul"ng service combining external experts and IP protec"on
                  mechanisms to expedite solu"ons to major corporate challenges
            "     Enterprise IT strategy Planning and contract support for enterprise IT leaders; produc"vity and infrastructure
                  agility benchmarking and best prac"ce
            "     IT vendor strategy and marke!ng Customer analysis and deep research for IT vendors targe"ng mid-market and
                  enterprise markets
            "     Sustainability strategy Modelling and benchmarking for organisa"ons wishing to embed sustainable prac"ces and
                  mi"gate strategic risks

        Research and benchmarking
            "     Research Pla"orm A exible so$ware pla#orm for organisa"ons to deploy their own surveys and benchmarks,
                  both internal (e.g. employee surveys), or external (e.g. market intelligence)
            "     Media Pla"orm A research pla#orm for online media customers, providing an engagement environment to
                  encourage reader registra"on and generate new revenues

        Sales enablement
            "     The Customer Insight Pla"orm A so$ware pla#orm that supports deployment of sales guidance, customer
                  assessments, marke"ng collaterals, ROI tools, etc. direct to a seller’s laptop for online and o'ine opera"on; for
                  both direct and channel sales support.


        The Bathwick Group also includes the ThinkAgain Partnership LLP, a global collabora"ve research network, which brings
        together academics, writers, business and poli"cal leaders to generate new insights into business produc"vity and perform-
        ance, geo-poli"cal and environmental issues.




                                                                                                                           LVL03001-USEN-00
All contents © The Bathwick Group Ltd 2010                             14                                                     www.bathwick.com

Weitere ähnliche Inhalte

Was ist angesagt?

The Economics of the Cloud
The Economics of the CloudThe Economics of the Cloud
The Economics of the Cloudrtharms
 
The Economics Of The Cloud
The Economics Of The CloudThe Economics Of The Cloud
The Economics Of The Cloudjessiethe3rd
 
The sky's the limit
The sky's the limitThe sky's the limit
The sky's the limitIvan Zoratti
 
[OSDC.tw 2011] The Path to Pass into PaaS -- How We Build the Solution
[OSDC.tw 2011] The Path to Pass into PaaS -- How We Build the Solution[OSDC.tw 2011] The Path to Pass into PaaS -- How We Build the Solution
[OSDC.tw 2011] The Path to Pass into PaaS -- How We Build the SolutionJeff Hung
 
Cloud Economics Cloud Connect 2011 Rolf Harms
Cloud Economics Cloud Connect 2011 Rolf HarmsCloud Economics Cloud Connect 2011 Rolf Harms
Cloud Economics Cloud Connect 2011 Rolf Harmsrtharms
 
Building a foundation for a high-performance, low cost private cloud
Building a foundation for a  high-performance, low cost  private cloudBuilding a foundation for a  high-performance, low cost  private cloud
Building a foundation for a high-performance, low cost private cloudIBM India Smarter Computing
 
Assessing no sql databases for telecom applications
Assessing no sql databases for telecom applicationsAssessing no sql databases for telecom applications
Assessing no sql databases for telecom applicationsJoão Gabriel Lima
 
What's After Cloud Computing?
What's After Cloud Computing?What's After Cloud Computing?
What's After Cloud Computing?InnoTech
 
Towards the convergence of digital tv systems
Towards the convergence of digital tv systemsTowards the convergence of digital tv systems
Towards the convergence of digital tv systemsLILI
 
CCA09 Cloud Computing Standards and OCCI
CCA09 Cloud Computing Standards and OCCICCA09 Cloud Computing Standards and OCCI
CCA09 Cloud Computing Standards and OCCIbefreax
 
HPC Cloud Computing with OpenNebula
HPC Cloud Computing with OpenNebulaHPC Cloud Computing with OpenNebula
HPC Cloud Computing with OpenNebulaIgnacio M. Llorente
 
SOFTBANK TELECOM Corp.
SOFTBANK TELECOM Corp.SOFTBANK TELECOM Corp.
SOFTBANK TELECOM Corp.NetApp
 

Was ist angesagt? (14)

The Economics of the Cloud
The Economics of the CloudThe Economics of the Cloud
The Economics of the Cloud
 
The Economics Of The Cloud
The Economics Of The CloudThe Economics Of The Cloud
The Economics Of The Cloud
 
The sky's the limit
The sky's the limitThe sky's the limit
The sky's the limit
 
[OSDC.tw 2011] The Path to Pass into PaaS -- How We Build the Solution
[OSDC.tw 2011] The Path to Pass into PaaS -- How We Build the Solution[OSDC.tw 2011] The Path to Pass into PaaS -- How We Build the Solution
[OSDC.tw 2011] The Path to Pass into PaaS -- How We Build the Solution
 
Cloud Economics Cloud Connect 2011 Rolf Harms
Cloud Economics Cloud Connect 2011 Rolf HarmsCloud Economics Cloud Connect 2011 Rolf Harms
Cloud Economics Cloud Connect 2011 Rolf Harms
 
Building a foundation for a high-performance, low cost private cloud
Building a foundation for a  high-performance, low cost  private cloudBuilding a foundation for a  high-performance, low cost  private cloud
Building a foundation for a high-performance, low cost private cloud
 
Assessing no sql databases for telecom applications
Assessing no sql databases for telecom applicationsAssessing no sql databases for telecom applications
Assessing no sql databases for telecom applications
 
What's After Cloud Computing?
What's After Cloud Computing?What's After Cloud Computing?
What's After Cloud Computing?
 
Towards the convergence of digital tv systems
Towards the convergence of digital tv systemsTowards the convergence of digital tv systems
Towards the convergence of digital tv systems
 
CCA09 Cloud Computing Standards and OCCI
CCA09 Cloud Computing Standards and OCCICCA09 Cloud Computing Standards and OCCI
CCA09 Cloud Computing Standards and OCCI
 
Smals
Smals Smals
Smals
 
HPC Cloud Computing with OpenNebula
HPC Cloud Computing with OpenNebulaHPC Cloud Computing with OpenNebula
HPC Cloud Computing with OpenNebula
 
Euronics XIV client reference
Euronics XIV client referenceEuronics XIV client reference
Euronics XIV client reference
 
SOFTBANK TELECOM Corp.
SOFTBANK TELECOM Corp.SOFTBANK TELECOM Corp.
SOFTBANK TELECOM Corp.
 

Andere mochten auch

Arquitectura de hardware de equipo de computo
Arquitectura de hardware de equipo de computo Arquitectura de hardware de equipo de computo
Arquitectura de hardware de equipo de computo sofia0107
 
Triptic IES taradell
Triptic IES taradellTriptic IES taradell
Triptic IES taradellampa
 
цаг хожих арга
цаг хожих аргацаг хожих арга
цаг хожих аргаSoko_92
 
Libres De Violencia
Libres De ViolenciaLibres De Violencia
Libres De ViolenciaAlhejandro
 
одно прекрасное мгновение юрия гагарина
одно прекрасное мгновение юрия гагаринаодно прекрасное мгновение юрия гагарина
одно прекрасное мгновение юрия гагаринаAnatoly Glushenkov
 
English central user's guide
English central user's guideEnglish central user's guide
English central user's guidegenryoy
 
Dossier de presse carte interactive & appli
Dossier de presse carte interactive & appliDossier de presse carte interactive & appli
Dossier de presse carte interactive & appliOtbmo
 
Bwii Wimax Eureka Certificate
Bwii Wimax  Eureka CertificateBwii Wimax  Eureka Certificate
Bwii Wimax Eureka Certificatelbrambi
 
Tips to public speaking-how to entertain the audience from the stage
Tips to public speaking-how to entertain the audience from the stageTips to public speaking-how to entertain the audience from the stage
Tips to public speaking-how to entertain the audience from the stageZAINI ABDUL WAHAB
 
Jihad kashmir-dr-asmatullah
Jihad kashmir-dr-asmatullahJihad kashmir-dr-asmatullah
Jihad kashmir-dr-asmatullahMurabitoon
 
Lectura no. 1 planeacion y tipos de estategias
Lectura no. 1 planeacion y tipos de estategiasLectura no. 1 planeacion y tipos de estategias
Lectura no. 1 planeacion y tipos de estategiascamomejia
 
What are the conventions of an Opening Sequence?
What are the conventions of an Opening Sequence?What are the conventions of an Opening Sequence?
What are the conventions of an Opening Sequence?JessicaBlyth12cms2011
 
Publicación3 teresa baldiviezo
Publicación3 teresa baldiviezoPublicación3 teresa baldiviezo
Publicación3 teresa baldiviezoTaek Sarmiento
 
Think bike, think bikers advert
Think bike, think bikers advertThink bike, think bikers advert
Think bike, think bikers advertLouiseMaher18
 
Reconnect 2014 web final
Reconnect 2014 web finalReconnect 2014 web final
Reconnect 2014 web finalLori Matherson
 

Andere mochten auch (20)

Arquitectura de hardware de equipo de computo
Arquitectura de hardware de equipo de computo Arquitectura de hardware de equipo de computo
Arquitectura de hardware de equipo de computo
 
Triptic IES taradell
Triptic IES taradellTriptic IES taradell
Triptic IES taradell
 
цаг хожих арга
цаг хожих аргацаг хожих арга
цаг хожих арга
 
Libres De Violencia
Libres De ViolenciaLibres De Violencia
Libres De Violencia
 
одно прекрасное мгновение юрия гагарина
одно прекрасное мгновение юрия гагаринаодно прекрасное мгновение юрия гагарина
одно прекрасное мгновение юрия гагарина
 
English central user's guide
English central user's guideEnglish central user's guide
English central user's guide
 
Dossier de presse carte interactive & appli
Dossier de presse carte interactive & appliDossier de presse carte interactive & appli
Dossier de presse carte interactive & appli
 
Bwii Wimax Eureka Certificate
Bwii Wimax  Eureka CertificateBwii Wimax  Eureka Certificate
Bwii Wimax Eureka Certificate
 
Tips to public speaking-how to entertain the audience from the stage
Tips to public speaking-how to entertain the audience from the stageTips to public speaking-how to entertain the audience from the stage
Tips to public speaking-how to entertain the audience from the stage
 
Jihad kashmir-dr-asmatullah
Jihad kashmir-dr-asmatullahJihad kashmir-dr-asmatullah
Jihad kashmir-dr-asmatullah
 
Gto1014 td
Gto1014 tdGto1014 td
Gto1014 td
 
Lectura no. 1 planeacion y tipos de estategias
Lectura no. 1 planeacion y tipos de estategiasLectura no. 1 planeacion y tipos de estategias
Lectura no. 1 planeacion y tipos de estategias
 
What are the conventions of an Opening Sequence?
What are the conventions of an Opening Sequence?What are the conventions of an Opening Sequence?
What are the conventions of an Opening Sequence?
 
Publicación3 teresa baldiviezo
Publicación3 teresa baldiviezoPublicación3 teresa baldiviezo
Publicación3 teresa baldiviezo
 
Irish brands Newbridge silerware
Irish brands Newbridge silerwareIrish brands Newbridge silerware
Irish brands Newbridge silerware
 
Think bike, think bikers advert
Think bike, think bikers advertThink bike, think bikers advert
Think bike, think bikers advert
 
Puerto de ostia
Puerto de ostiaPuerto de ostia
Puerto de ostia
 
Teléfono telefoniño
Teléfono telefoniñoTeléfono telefoniño
Teléfono telefoniño
 
Reconnect 2014 web final
Reconnect 2014 web finalReconnect 2014 web final
Reconnect 2014 web final
 
Aprendizaje basado en proyectos
Aprendizaje  basado en  proyectosAprendizaje  basado en  proyectos
Aprendizaje basado en proyectos
 

Ähnlich wie The role of the mainframe as a host environment for enterprise cloud computing

Cloud deep-dive0212
Cloud deep-dive0212Cloud deep-dive0212
Cloud deep-dive0212Accenture
 
Cloud Computing And Virtualization
Cloud Computing And VirtualizationCloud Computing And Virtualization
Cloud Computing And VirtualizationSonali Parab
 
Regarding Clouds, Mainframes, and Desktops … and Linux
Regarding Clouds, Mainframes, and Desktops … and LinuxRegarding Clouds, Mainframes, and Desktops … and Linux
Regarding Clouds, Mainframes, and Desktops … and LinuxRobert Sutor
 
Say Goodbye to DIY Data Centers
Say Goodbye to DIY Data CentersSay Goodbye to DIY Data Centers
Say Goodbye to DIY Data CentersRackspace
 
Computing in the clouds weiss
Computing in the clouds weissComputing in the clouds weiss
Computing in the clouds weissthreesixty
 
Solar Powered MicroServers - Green Computing
Solar Powered MicroServers - Green ComputingSolar Powered MicroServers - Green Computing
Solar Powered MicroServers - Green ComputingPaul Morse
 
OpenStack Architected Like AWS (and GCP)
OpenStack Architected Like AWS (and GCP)OpenStack Architected Like AWS (and GCP)
OpenStack Architected Like AWS (and GCP)Randy Bias
 
Widespread Cloud Adoption: What's Taking So Long?
Widespread Cloud Adoption: What's Taking So Long?Widespread Cloud Adoption: What's Taking So Long?
Widespread Cloud Adoption: What's Taking So Long?Cartesian (formerly CSMG)
 
Cloud Computing Big Data Is Future Of It
Cloud Computing Big  Data Is Future Of ItCloud Computing Big  Data Is Future Of It
Cloud Computing Big Data Is Future Of ItAman Ghei
 
Datacenter Of The Future Article - - Jon Greaves 1
Datacenter Of The Future Article - -  Jon Greaves 1Datacenter Of The Future Article - -  Jon Greaves 1
Datacenter Of The Future Article - - Jon Greaves 1Bill Alatis
 
Past & Future Of The Cloud Nott Tues Nov 09
Past & Future Of The Cloud Nott Tues Nov 09Past & Future Of The Cloud Nott Tues Nov 09
Past & Future Of The Cloud Nott Tues Nov 09Nickbarker
 
Teja pp matter
Teja pp matterTeja pp matter
Teja pp matter9505567198
 
Changing Landscape of Data Centers
Changing Landscape of Data CentersChanging Landscape of Data Centers
Changing Landscape of Data CentersSuhas Kelkar
 
Deconstructing the brian paradox
Deconstructing the brian paradoxDeconstructing the brian paradox
Deconstructing the brian paradoxChetan Venkatesh
 
Cloud computingworldforum
Cloud computingworldforumCloud computingworldforum
Cloud computingworldforumCharles Mok
 
Creating the open_cloud_wp_web2
Creating the open_cloud_wp_web2Creating the open_cloud_wp_web2
Creating the open_cloud_wp_web2Srikanth Minnam
 
Creating Cloud Communities
Creating Cloud CommunitiesCreating Cloud Communities
Creating Cloud CommunitiesPeter Coffee
 

Ähnlich wie The role of the mainframe as a host environment for enterprise cloud computing (20)

Cloud deep-dive0212
Cloud deep-dive0212Cloud deep-dive0212
Cloud deep-dive0212
 
Cloud Computing
Cloud ComputingCloud Computing
Cloud Computing
 
Cloud Computing And Virtualization
Cloud Computing And VirtualizationCloud Computing And Virtualization
Cloud Computing And Virtualization
 
Regarding Clouds, Mainframes, and Desktops … and Linux
Regarding Clouds, Mainframes, and Desktops … and LinuxRegarding Clouds, Mainframes, and Desktops … and Linux
Regarding Clouds, Mainframes, and Desktops … and Linux
 
Say Goodbye to DIY Data Centers
Say Goodbye to DIY Data CentersSay Goodbye to DIY Data Centers
Say Goodbye to DIY Data Centers
 
The Enterprise Cloud
The Enterprise CloudThe Enterprise Cloud
The Enterprise Cloud
 
Computing in the clouds weiss
Computing in the clouds weissComputing in the clouds weiss
Computing in the clouds weiss
 
Solar Powered MicroServers - Green Computing
Solar Powered MicroServers - Green ComputingSolar Powered MicroServers - Green Computing
Solar Powered MicroServers - Green Computing
 
OpenStack Architected Like AWS (and GCP)
OpenStack Architected Like AWS (and GCP)OpenStack Architected Like AWS (and GCP)
OpenStack Architected Like AWS (and GCP)
 
Cloud computing
Cloud computingCloud computing
Cloud computing
 
Widespread Cloud Adoption: What's Taking So Long?
Widespread Cloud Adoption: What's Taking So Long?Widespread Cloud Adoption: What's Taking So Long?
Widespread Cloud Adoption: What's Taking So Long?
 
Cloud Computing Big Data Is Future Of It
Cloud Computing Big  Data Is Future Of ItCloud Computing Big  Data Is Future Of It
Cloud Computing Big Data Is Future Of It
 
Datacenter Of The Future Article - - Jon Greaves 1
Datacenter Of The Future Article - -  Jon Greaves 1Datacenter Of The Future Article - -  Jon Greaves 1
Datacenter Of The Future Article - - Jon Greaves 1
 
Past & Future Of The Cloud Nott Tues Nov 09
Past & Future Of The Cloud Nott Tues Nov 09Past & Future Of The Cloud Nott Tues Nov 09
Past & Future Of The Cloud Nott Tues Nov 09
 
Teja pp matter
Teja pp matterTeja pp matter
Teja pp matter
 
Changing Landscape of Data Centers
Changing Landscape of Data CentersChanging Landscape of Data Centers
Changing Landscape of Data Centers
 
Deconstructing the brian paradox
Deconstructing the brian paradoxDeconstructing the brian paradox
Deconstructing the brian paradox
 
Cloud computingworldforum
Cloud computingworldforumCloud computingworldforum
Cloud computingworldforum
 
Creating the open_cloud_wp_web2
Creating the open_cloud_wp_web2Creating the open_cloud_wp_web2
Creating the open_cloud_wp_web2
 
Creating Cloud Communities
Creating Cloud CommunitiesCreating Cloud Communities
Creating Cloud Communities
 

Mehr von IBM India Smarter Computing

Using the IBM XIV Storage System in OpenStack Cloud Environments
Using the IBM XIV Storage System in OpenStack Cloud Environments Using the IBM XIV Storage System in OpenStack Cloud Environments
Using the IBM XIV Storage System in OpenStack Cloud Environments IBM India Smarter Computing
 
TSL03104USEN Exploring VMware vSphere Storage API for Array Integration on th...
TSL03104USEN Exploring VMware vSphere Storage API for Array Integration on th...TSL03104USEN Exploring VMware vSphere Storage API for Array Integration on th...
TSL03104USEN Exploring VMware vSphere Storage API for Array Integration on th...IBM India Smarter Computing
 
A Comparison of PowerVM and Vmware Virtualization Performance
A Comparison of PowerVM and Vmware Virtualization PerformanceA Comparison of PowerVM and Vmware Virtualization Performance
A Comparison of PowerVM and Vmware Virtualization PerformanceIBM India Smarter Computing
 
IBM pureflex system and vmware vcloud enterprise suite reference architecture
IBM pureflex system and vmware vcloud enterprise suite reference architectureIBM pureflex system and vmware vcloud enterprise suite reference architecture
IBM pureflex system and vmware vcloud enterprise suite reference architectureIBM India Smarter Computing
 

Mehr von IBM India Smarter Computing (20)

Using the IBM XIV Storage System in OpenStack Cloud Environments
Using the IBM XIV Storage System in OpenStack Cloud Environments Using the IBM XIV Storage System in OpenStack Cloud Environments
Using the IBM XIV Storage System in OpenStack Cloud Environments
 
All-flash Needs End to End Storage Efficiency
All-flash Needs End to End Storage EfficiencyAll-flash Needs End to End Storage Efficiency
All-flash Needs End to End Storage Efficiency
 
TSL03104USEN Exploring VMware vSphere Storage API for Array Integration on th...
TSL03104USEN Exploring VMware vSphere Storage API for Array Integration on th...TSL03104USEN Exploring VMware vSphere Storage API for Array Integration on th...
TSL03104USEN Exploring VMware vSphere Storage API for Array Integration on th...
 
IBM FlashSystem 840 Product Guide
IBM FlashSystem 840 Product GuideIBM FlashSystem 840 Product Guide
IBM FlashSystem 840 Product Guide
 
IBM System x3250 M5
IBM System x3250 M5IBM System x3250 M5
IBM System x3250 M5
 
IBM NeXtScale nx360 M4
IBM NeXtScale nx360 M4IBM NeXtScale nx360 M4
IBM NeXtScale nx360 M4
 
IBM System x3650 M4 HD
IBM System x3650 M4 HDIBM System x3650 M4 HD
IBM System x3650 M4 HD
 
IBM System x3300 M4
IBM System x3300 M4IBM System x3300 M4
IBM System x3300 M4
 
IBM System x iDataPlex dx360 M4
IBM System x iDataPlex dx360 M4IBM System x iDataPlex dx360 M4
IBM System x iDataPlex dx360 M4
 
IBM System x3500 M4
IBM System x3500 M4IBM System x3500 M4
IBM System x3500 M4
 
IBM System x3550 M4
IBM System x3550 M4IBM System x3550 M4
IBM System x3550 M4
 
IBM System x3650 M4
IBM System x3650 M4IBM System x3650 M4
IBM System x3650 M4
 
IBM System x3500 M3
IBM System x3500 M3IBM System x3500 M3
IBM System x3500 M3
 
IBM System x3400 M3
IBM System x3400 M3IBM System x3400 M3
IBM System x3400 M3
 
IBM System x3250 M3
IBM System x3250 M3IBM System x3250 M3
IBM System x3250 M3
 
IBM System x3200 M3
IBM System x3200 M3IBM System x3200 M3
IBM System x3200 M3
 
IBM PowerVC Introduction and Configuration
IBM PowerVC Introduction and ConfigurationIBM PowerVC Introduction and Configuration
IBM PowerVC Introduction and Configuration
 
A Comparison of PowerVM and Vmware Virtualization Performance
A Comparison of PowerVM and Vmware Virtualization PerformanceA Comparison of PowerVM and Vmware Virtualization Performance
A Comparison of PowerVM and Vmware Virtualization Performance
 
IBM pureflex system and vmware vcloud enterprise suite reference architecture
IBM pureflex system and vmware vcloud enterprise suite reference architectureIBM pureflex system and vmware vcloud enterprise suite reference architecture
IBM pureflex system and vmware vcloud enterprise suite reference architecture
 
X6: The sixth generation of EXA Technology
X6: The sixth generation of EXA TechnologyX6: The sixth generation of EXA Technology
X6: The sixth generation of EXA Technology
 

Kürzlich hochgeladen

Meet the new FSP 3000 M-Flex800™
Meet the new FSP 3000 M-Flex800™Meet the new FSP 3000 M-Flex800™
Meet the new FSP 3000 M-Flex800™Adtran
 
IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019
IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019
IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019IES VE
 
Computer 10: Lesson 10 - Online Crimes and Hazards
Computer 10: Lesson 10 - Online Crimes and HazardsComputer 10: Lesson 10 - Online Crimes and Hazards
Computer 10: Lesson 10 - Online Crimes and HazardsSeth Reyes
 
Linked Data in Production: Moving Beyond Ontologies
Linked Data in Production: Moving Beyond OntologiesLinked Data in Production: Moving Beyond Ontologies
Linked Data in Production: Moving Beyond OntologiesDavid Newbury
 
GenAI and AI GCC State of AI_Object Automation Inc
GenAI and AI GCC State of AI_Object Automation IncGenAI and AI GCC State of AI_Object Automation Inc
GenAI and AI GCC State of AI_Object Automation IncObject Automation
 
Cybersecurity Workshop #1.pptx
Cybersecurity Workshop #1.pptxCybersecurity Workshop #1.pptx
Cybersecurity Workshop #1.pptxGDSC PJATK
 
Nanopower In Semiconductor Industry.pdf
Nanopower  In Semiconductor Industry.pdfNanopower  In Semiconductor Industry.pdf
Nanopower In Semiconductor Industry.pdfPedro Manuel
 
Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...
Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...
Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...DianaGray10
 
OpenShift Commons Paris - Choose Your Own Observability Adventure
OpenShift Commons Paris - Choose Your Own Observability AdventureOpenShift Commons Paris - Choose Your Own Observability Adventure
OpenShift Commons Paris - Choose Your Own Observability AdventureEric D. Schabell
 
AI Fame Rush Review – Virtual Influencer Creation In Just Minutes
AI Fame Rush Review – Virtual Influencer Creation In Just MinutesAI Fame Rush Review – Virtual Influencer Creation In Just Minutes
AI Fame Rush Review – Virtual Influencer Creation In Just MinutesMd Hossain Ali
 
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCost
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCostKubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCost
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCostMatt Ray
 
UiPath Studio Web workshop series - Day 7
UiPath Studio Web workshop series - Day 7UiPath Studio Web workshop series - Day 7
UiPath Studio Web workshop series - Day 7DianaGray10
 
UiPath Studio Web workshop series - Day 6
UiPath Studio Web workshop series - Day 6UiPath Studio Web workshop series - Day 6
UiPath Studio Web workshop series - Day 6DianaGray10
 
Videogame localization & technology_ how to enhance the power of translation.pdf
Videogame localization & technology_ how to enhance the power of translation.pdfVideogame localization & technology_ how to enhance the power of translation.pdf
Videogame localization & technology_ how to enhance the power of translation.pdfinfogdgmi
 
PicPay - GenAI Finance Assistant - ChatGPT for Customer Service
PicPay - GenAI Finance Assistant - ChatGPT for Customer ServicePicPay - GenAI Finance Assistant - ChatGPT for Customer Service
PicPay - GenAI Finance Assistant - ChatGPT for Customer ServiceRenan Moreira de Oliveira
 
UiPath Community: AI for UiPath Automation Developers
UiPath Community: AI for UiPath Automation DevelopersUiPath Community: AI for UiPath Automation Developers
UiPath Community: AI for UiPath Automation DevelopersUiPathCommunity
 
COMPUTER 10: Lesson 7 - File Storage and Online Collaboration
COMPUTER 10: Lesson 7 - File Storage and Online CollaborationCOMPUTER 10: Lesson 7 - File Storage and Online Collaboration
COMPUTER 10: Lesson 7 - File Storage and Online Collaborationbruanjhuli
 
Bird eye's view on Camunda open source ecosystem
Bird eye's view on Camunda open source ecosystemBird eye's view on Camunda open source ecosystem
Bird eye's view on Camunda open source ecosystemAsko Soukka
 
IaC & GitOps in a Nutshell - a FridayInANuthshell Episode.pdf
IaC & GitOps in a Nutshell - a FridayInANuthshell Episode.pdfIaC & GitOps in a Nutshell - a FridayInANuthshell Episode.pdf
IaC & GitOps in a Nutshell - a FridayInANuthshell Episode.pdfDaniel Santiago Silva Capera
 
Using IESVE for Loads, Sizing and Heat Pump Modeling to Achieve Decarbonization
Using IESVE for Loads, Sizing and Heat Pump Modeling to Achieve DecarbonizationUsing IESVE for Loads, Sizing and Heat Pump Modeling to Achieve Decarbonization
Using IESVE for Loads, Sizing and Heat Pump Modeling to Achieve DecarbonizationIES VE
 

Kürzlich hochgeladen (20)

Meet the new FSP 3000 M-Flex800™
Meet the new FSP 3000 M-Flex800™Meet the new FSP 3000 M-Flex800™
Meet the new FSP 3000 M-Flex800™
 
IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019
IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019
IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019
 
Computer 10: Lesson 10 - Online Crimes and Hazards
Computer 10: Lesson 10 - Online Crimes and HazardsComputer 10: Lesson 10 - Online Crimes and Hazards
Computer 10: Lesson 10 - Online Crimes and Hazards
 
Linked Data in Production: Moving Beyond Ontologies
Linked Data in Production: Moving Beyond OntologiesLinked Data in Production: Moving Beyond Ontologies
Linked Data in Production: Moving Beyond Ontologies
 
GenAI and AI GCC State of AI_Object Automation Inc
GenAI and AI GCC State of AI_Object Automation IncGenAI and AI GCC State of AI_Object Automation Inc
GenAI and AI GCC State of AI_Object Automation Inc
 
Cybersecurity Workshop #1.pptx
Cybersecurity Workshop #1.pptxCybersecurity Workshop #1.pptx
Cybersecurity Workshop #1.pptx
 
Nanopower In Semiconductor Industry.pdf
Nanopower  In Semiconductor Industry.pdfNanopower  In Semiconductor Industry.pdf
Nanopower In Semiconductor Industry.pdf
 
Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...
Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...
Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...
 
OpenShift Commons Paris - Choose Your Own Observability Adventure
OpenShift Commons Paris - Choose Your Own Observability AdventureOpenShift Commons Paris - Choose Your Own Observability Adventure
OpenShift Commons Paris - Choose Your Own Observability Adventure
 
AI Fame Rush Review – Virtual Influencer Creation In Just Minutes
AI Fame Rush Review – Virtual Influencer Creation In Just MinutesAI Fame Rush Review – Virtual Influencer Creation In Just Minutes
AI Fame Rush Review – Virtual Influencer Creation In Just Minutes
 
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCost
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCostKubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCost
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCost
 
UiPath Studio Web workshop series - Day 7
UiPath Studio Web workshop series - Day 7UiPath Studio Web workshop series - Day 7
UiPath Studio Web workshop series - Day 7
 
UiPath Studio Web workshop series - Day 6
UiPath Studio Web workshop series - Day 6UiPath Studio Web workshop series - Day 6
UiPath Studio Web workshop series - Day 6
 
Videogame localization & technology_ how to enhance the power of translation.pdf
Videogame localization & technology_ how to enhance the power of translation.pdfVideogame localization & technology_ how to enhance the power of translation.pdf
Videogame localization & technology_ how to enhance the power of translation.pdf
 
PicPay - GenAI Finance Assistant - ChatGPT for Customer Service
PicPay - GenAI Finance Assistant - ChatGPT for Customer ServicePicPay - GenAI Finance Assistant - ChatGPT for Customer Service
PicPay - GenAI Finance Assistant - ChatGPT for Customer Service
 
UiPath Community: AI for UiPath Automation Developers
UiPath Community: AI for UiPath Automation DevelopersUiPath Community: AI for UiPath Automation Developers
UiPath Community: AI for UiPath Automation Developers
 
COMPUTER 10: Lesson 7 - File Storage and Online Collaboration
COMPUTER 10: Lesson 7 - File Storage and Online CollaborationCOMPUTER 10: Lesson 7 - File Storage and Online Collaboration
COMPUTER 10: Lesson 7 - File Storage and Online Collaboration
 
Bird eye's view on Camunda open source ecosystem
Bird eye's view on Camunda open source ecosystemBird eye's view on Camunda open source ecosystem
Bird eye's view on Camunda open source ecosystem
 
IaC & GitOps in a Nutshell - a FridayInANuthshell Episode.pdf
IaC & GitOps in a Nutshell - a FridayInANuthshell Episode.pdfIaC & GitOps in a Nutshell - a FridayInANuthshell Episode.pdf
IaC & GitOps in a Nutshell - a FridayInANuthshell Episode.pdf
 
Using IESVE for Loads, Sizing and Heat Pump Modeling to Achieve Decarbonization
Using IESVE for Loads, Sizing and Heat Pump Modeling to Achieve DecarbonizationUsing IESVE for Loads, Sizing and Heat Pump Modeling to Achieve Decarbonization
Using IESVE for Loads, Sizing and Heat Pump Modeling to Achieve Decarbonization
 

The role of the mainframe as a host environment for enterprise cloud computing

  • 1. TheBathwickGroup A Bathwick Group white paper The mainframe and the cloud The role of the mainframe as a host environment for enterprise cloud computing Document Number: BC-INF-W-00068UK-EN-00 All contents © The Bathwick Group Ltd 2010 Author: Gary Barne! Published: November 2010
  • 2. TheBathwickGroup The mainframe and the cloud - November 2010 Abstract Nearly all of the discussion and debate surrounding cloud compu!ng presumes that the ideal enterprise cloud environment will be built on x86 processors in a massively scaled-out architecture. Un!l recently there has been very li"le discussion about the poten!al benets that the mainframe o#ers the enterprise as a host for a cloud compu!ng environment. This paper presents the poten!al benets that enterprises could derive from the mainframe as a cloud environment in terms of cost, reliability, scalability, security and exibility. We also set out to highlight the changes that have taken place within the mainframe eco-system that mean many of the tradi!onal cri!cisms of the mainframe pla$orm are simply out of date. Key facts The latest version of the mainframe introduces: ! " A consolida"on pla#orm that could reduce energy consump"on, oor space required, labour costs and so$ware licensing costs by as much as 90% ! " A virtualisa"on environment that can support 1000s of mainframe, Linux, and Java environments, all within one physical machine ! " An environment within which System p blades (running AIX) and – in the rst half of 2011 - Linux running on System x blades can run directly alongside the mainframe using the same virtual network and bene%ng from similar levels of reliability, availability and scalability as the mainframe itself. Key messages ! " Contrary to popular myth, the mainframe is alive and well. We o$en hear commentators claiming that the era of the mainframe is “over”, despite the fact that IBM’s mainframe franchise con"nues to grow. In one sense, these statements are true; the mainframe of the 1980s is dead. The 80s mainframe has been replaced by a pla#orm that is more open, more cost e&ec"ve and more exible than ever before. ! " The mainframe represents a compelling proposi!on as a cloud hos!ng environment. The mainframe o&ers an alterna"ve to the massively scaled-out environments that are currently being touted as the future of cloud. Scale-out architectures undoubtedly bring advantages, and in some cases they certainly do represent the best approach, but the scale-up architecture of the mainframe brings its own benets that mean that for some workloads it o&ers signicant advantages over scaled-out systems. The key lies in understanding which situa"ons indicate one approach or the other. "! Dispelling the myths of complexity, integra!on and cost. Cri"cs of the mainframe typically cite the same set of problems with the mainframe pla#orm – “It is expensive”, “It is a closed environment”, “It is hard to develop so$ware for the mainframe”. These cri"cisms may have been true 15 years ago but in the intervening "me the mainframe has changed radically. All contents © The Bathwick Group Ltd 2010 2 www.bathwick.com
  • 3. TheBathwickGroup The mainframe and the cloud - November 2010 "! IBM’s latest mainframe transforms the value proposi!on. The mainframe value proposi"on has been undergoing a con"nuous transforma"on over the past two decades in terms of the so$ware stack and the hardware. The latest incarna"on of the mainframe, the zEnterprise, takes this evolu"on a signicant way further by making it possible to combine mainframe and non- mainframe hardware within a single managed environment. "! If you’re considering consolida!ng SAP, your web infrastructure or your BI pla"orms, consider the mainframe. IBM has focussed on exactly these applica"on scenarios in developing the new mainframe and even organisa"ons that don’t currently have a mainframe might nd the overall value proposi"on compelling. Dispelling some mainframe myths The next "me someone says that the mainframe is dead or that the era of the mainframe is over, ask them what compu"ng pla#orm manages their credit-card payments, controls the ATM network they use to withdraw cash, or underpins the globe’s airline "cke"ng systems; almost without excep"on it’s a mainframe. In the 1980s the mainframe was perceived as costly, complex, and closed. There was some jus"ca"on for this view and it helped to drive the adop"on of client-server compu"ng as a cheaper, less complex and more open approach to compu"ng. However, over the past two decades, the mainframe has undergone a drama"c transforma"on aimed at directly addressing the shortcomings of the 80s mainframe. The mainframe of the 1980s is dead; welcome to the mainframe of the 2010s In the 1980s client-server compu"ng emerged as a response to the cost and complexity associated with the mainframe. Client-server compu"ng promised lower cost, greater exibility, and an environment that was far more open in terms of the ease with which di&erent systems could be integrated with one another. When you look back at the mainframe of the 1980s it’s easy to see why client-server compu"ng was regarded as a be!er, less expensive alterna"ve. 1980s – Closed, proprietary, costly Mainframe Distributed Figure 1. The mainframe of the 1980s All contents © The Bathwick Group Ltd 2010 3 www.bathwick.com
  • 4. TheBathwickGroup The mainframe and the cloud - November 2010 In the 1980s the mainframe was a fortress. On one hand, it was secure, reliable and scalable, but on the other hand it was expensive and inaccessible. Companies that wanted to integrate mainframe data with their emerging client-server applica"ons o$en had to resort to ETL solu"ons that dumped the mainframe data onto tape so that it could then be uploaded into the client-server system. Within the mainframe environment, you were free to develop your own so$ware – provided you used mainframe programming languages like Fortran, COBOL or PL/1. Meanwhile, within the world of distributed compu"ng there was an explosion of innova"on in the world of programming languages, development tools and middleware. 1990s – Fortress mainframe gets a drawbridge Mainframe Distributed Figure 2. The mainframe of the 1990s In the 1990s IBM took the rst serious steps to make it easier to integrate mainframe-based applica"ons and non-mainframe applica"ons by introducing connec"vity middleware and a number of third-par"es (including the likes of Tuxedo and So$ware AG) “ I predict that the last main- developed alterna"ve middleware products that made it possible frame will be unplugged on (albeit at considerable expense and e&ort) to link the mainframe March 15, 1996. “ and non-mainframe worlds. Stewart Alsop, Infoworld March 1991 This was undoubtedly “progress”, but the mainframe was s"ll an ocean away from being a genuinely rst class ci"zen in the world of distributed compu"ng. Developers s"ll had to use “mainframe” development tools and programming languages. “Mainframe” middleware and the mainframe was s"ll a complex and costly environment to manage and run. The apparent failure of the mainframe to respond to the challenges presented by client-server compu"ng coupled with the sheer volume of innova"on and crea"vity within the client-server world resulted in a general belief that the mainframe was a dying pla#orm. This belief prompted one commentator to declare in 1991 that “The last mainframe will be unplugged on March 15, 1996”. All contents © The Bathwick Group Ltd 2010 4 www.bathwick.com
  • 5. TheBathwickGroup The mainframe and the cloud - November 2010 2000s – Reinven!ng the mainframe Distributed Mainframe Figure 3. The mainframe of the 2000s By the end of the 1990s, it was clear that IBM needed to take ac"on to re-invigorate the mainframe. In 1999 IBM announced plans to work with RedHat to bring Linux to the mainframe pla#orm and was able to o&er mainframe Linux as a standard op"on in 2000. A    Y The 2000s saw a number of profound changes to the mainframe environment. The most important theme surrounding these changes was the decision to take a systems oriented approach to the ongoing development of the mainframe by integra"ng the e&orts of IBM’s so$ware and hardware groups. In 2003 IBM made a very public commitment to the mainframe by introducing the Mainframe Charter – a promise to its customers that IBM would con"nue to invest in the pla#orm by inves"ng in mainframe hardware design, the so$ware ecosystem and in making it easier to manage and integrate the pla#orm within a distributed compu"ng environment. Throughout the 2000s, major mainframe hardware announcements went hand in hand with so$ware announcements and the por#olio of mainframe so$ware products became increasingly aligned with IBM’s non-mainframe so$ware por#olio. The improvement this strategy brought about in terms of integra"ng IBM’s di&erent mainframe o&erings and integra"on between IBM’s non-mainframe products as well was signicant for two key reasons; it reduced the cost and complexity associated with provisioning a mainframe and helped IBM to o&er a more integrated alterna"ve to the host of third-party products that are available for the mainframe from vendors like CA and Compuware. All contents © The Bathwick Group Ltd 2010 5 www.bathwick.com
  • 6. TheBathwickGroup The mainframe and the cloud - November 2010 So#ware and hardware alignment Throughout the 2000s IBM set about aligning its distributed and mainframe so$ware o&erings; ! "! CICS was enhanced to support web services with the introduc"on of SOAP for CICS ! "! IMS added support for RSS ! "! IBM’s mainframe development tools por#olio was integrated within the Ra"onal development tools family (Ra"onal Developer for System z) ! "! IBM’s systems management so$ware Tivoli was integrated so that key components (like iden"ty management, for example) could run on either distributed or mainframe systems ! "! DB2 was enhanced to improve the level of integra"on between the mainframe and non- mainframe versions of the database ! "! Mainframe management was drama"cally simplied with the release of Systems Director (a graphical management tool for the mainframe that automates a host of administra"on tasks) The hardware con"nued to evolve with improvements to the price/performance of the machine itself. This was further enhanced with the introduc"on of the ability to o'oad specic workloads (notably Java and Linux) from the main processors onto less specialist processors that were less costly themselves and also reduced the number of “core” MIPS that had to be licensed. At the same "me key IBM so$ware products – DB2, Ra"onal development tools, the Websphere middleware family and the Tivoli systems management suite all became increasingly aligned with both the mainframe and distributed pla#orms. SOA     One of the key changes in the mainframe was the addi"on of much improved support for SOA. Venerable technologies like CICS and IMS were all extended to make it possible to integrate them within an SOA-based environment. CICS now ships with support for SOAP and web services. Technologies like MQ Series act as a SOA bridge between the mainframe and other environments and the combina"on of middleware, security technology and management tools mean that for some organisa"ons the mainframe can serve as a single organisa"on-wide ESB. All contents © The Bathwick Group Ltd 2010 6 www.bathwick.com
  • 7. TheBathwickGroup The mainframe and the cloud - November 2010 2010 – The next genera!on Distributed Mainframe Figure 4. The next genera!on of the IBM mainframe In July 2010, IBM launched the zEnterprise server, which delivered two sets of benets. Firstly, the new mainframe delivers signicant improvements to the basic performance of the mainframe: "! The fastest general purpose CPU ever, running at 5.2Ghz "! Up to 40% performance improvement for tradi"onal mainframe workloads "! 80% improvement in Linux price/performance "! A consolida"on pla#orm that could reduce energy consump"on, oor space required, labour costs and so$ware licensing costs by as much as 90% "! Introduc"on of RAIM (redundant array of independent memory) which signicantly improves the reliability of system memory Secondly, and perhaps the most signicant feature of this release, is the integra"on the pla#orm o&ers with non-mainframe hardware. The zEnterprise can be a!ached to a zBX, which provides housing for up to 112 p-Series blades, running AIX; x-Series blades (running Linux) will be supported in the rst half of 2011. The zBX is connected to the zEnterprise machine via two private network connec"ons: a 10Gb Ethernet connec"on is used by the applica"ons to provide ultra highspeed connec"vity, and a second management network (running at 1Gb/s) is used by the machine’s management so$ware to control the di&erent hardware devices within the system. The blades housed within the zBX are managed, along with the mainframe itself, by a single management component called zManager. The zManager adds features to blade technology management that have tradi"onally been the sole preserve of the mainframe. The management component automa"cally discovers new hardware, manages and monitors running hardware, and supports running upgrades to so$ware and rmware. The zBX is built with levels of hardware redundancy comparable to the mainframe itself, so key components like power supplies, power distribu"on units, network switches and fans are all shipped with at least one addi"onal or redundant component. All contents © The Bathwick Group Ltd 2010 7 www.bathwick.com
  • 8. TheBathwickGroup The mainframe and the cloud - November 2010 Cloud in a box : Using the mainframe as the host for a cloud environment Cloud compu"ng is being heralded as the next evolu"on of systems architecture. The benets of cloud compu"ng are primarily built on the no"on that cloud environments allow you to gain benets in terms of cost and exibility by virtualizing the di&erent workloads and deploying them within an environment that can automa"cally scale-up or down as workload demands change over "me. Broadly speaking, cloud compu"ng environments can be; "! Public The environment is run on shared infrastructure accessible via the internet. In this scenario you pay a third party like Amazon (EC2) to run your workload and are charged according to the CPU "me your applica"ons use, the bandwidth they consume and the storage you require. "! Semi-Private The environment is run by a third party in much the same way as a tradi"onal outsourcing contract, but you are charged on the basis of u"lisa"on rather than paying a at fee. "! Private The environment is run on your own premises. In this scenario you maintain and run your own cloud infrastructure. C    While there are undoubtedly cases where public cloud is well suited to enterprise compu"ng, these cases tend to be where the required func"onality is well dened and has clear boundaries. IBM and others already o&er cloud-based development environments, for example, which allow development teams to quickly ramp up (or down) as their needs change. Public cloud is less suited to complex applica"ons where some of the processing or data can’t be run in the cloud. Imagine you have an ERP system running on your own infrastructure, and you want to deploy a web applica"on that is available to your customers. It is likely that you won’t be able to simply deploy your ERP applica"on in the cloud, so that will have to remain on your infrastructure. If you deploy the “web components” of the solu"on in a public cloud environment, you will have to manage the link between the cloud-based applica"on and your ERP system. While it is possible to do this, and you can do this securely by crea"ng a VPN that reaches beyond your infrastructure into the cloud environment, there is a real likelihood that you’ll encounter signicant performance headaches as transac"ons have to cross from the cloud environment, over the network, and into your infrastructure. If you could derive the same benets of scalability, ease of management, and automa"on by deploying cloud technology on your own infrastructure, you can avoid the problems that network latency creates – or at least you’ll be in a much be!er posi"on to manage them, as the whole applica"on will be running on your network. All contents © The Bathwick Group Ltd 2010 8 www.bathwick.com
  • 9. TheBathwickGroup The mainframe and the cloud - November 2010 Building your own cloud Many pundits will say that the only way to build a private cloud is to deploy a virtualisa"on technology, like VMware, and install a number of blades that will provide the processing horsepower for your cloud. But there is an alterna"ve to the conven"onal belief that “scale-out” is the way to deliver cloud compu"ng. Scale-up Scale-out Figure 5. Scale-up and Scale-out compu!ng Figure 5 shows, in very simple terms, the architectural di&erence between mainframe or scale-up compu"ng (where many images run within a single mul"-processor machine) and scale-out where one or more images run in a number of machines, that are then managed as a single en"ty by a virtualisa"on solu"on like VMware. Looking at the diagram above, it is worth asking the following ques"ons about the two approaches; "! What is the likely di&erence in physical footprint? "! What is the likely di&erence in power consump"on? "! How quickly can you increase capacity? "! What is the likely di&erence in reliability? "! What is the likely di&erence in the "me it would take for a request to be sent from one virtual image to another? Few proponents of scale-out compu"ng would claim that you can get more processing power per square foot of oor space by taking a scale-out approach; the total power consump"on (factoring in all power u"lised in order to operate the hardware – including cooling) per processor of a scale-out solu"on vs scale-up can be as much as 20 "mes more. Increasing capacity in a scale-out solu"on is o$en described as simply a case of “pushing another blade into a slot”, but that assumes that you have a blade that is wai"ng to be installed (which seems like a bit of a waste of capital and capacity) and that you have a vacant slot into which the new blade can be inserted. If you don’t have a blade standing by, or if you need to install a new blade enclosure All contents © The Bathwick Group Ltd 2010 9 www.bathwick.com
  • 10. TheBathwickGroup The mainframe and the cloud - November 2010 then adding capacity could take a considerable amount of "me. In the scale-up world, IBM can ship servers which are over-provisioned in terms of hardware, with the addi"onal processors ready to be enabled on- demand. Even if new hardware needs to be added, this can be requested as part of the standard procurement process and executed as part of the standard maintenance process. The ques"on of reliability is a complex one. Certainly, a catastrophic failure within an SMP based machine will result in a complete outage, while a catastrophic failure within a blade (or even a blade enclosure) may only result in a par"al outage for a subset of users. But it isn’t as simple as that. To begin with, the resilience of a scale-out applica"on to component failures depends, to a large degree, on the design of the applica"on. Some applica"ons have been designed to recover from a par"al failure while others may fail completely. In order to deliver high levels of reliability and availability in a scale-out environment, you need to specify mul"ple network devices, redundant power supplies, and middleware that supports fail-over and recovery. In the mainframe environment the same steps have been taken, but these are internalised within the machine itself. Taking a simple example like RAM (faulty RAM is one of the more common reasons for the failure of a blade) mainframe RAM is designed to be fault tolerant, faulty memory loca"ons can be quaran"ned and errors can be reported before they result in a system failure. With the latest release of the mainframe IBM has introduced RAIM – the RAM equivalent of RAID. The last ques"on depends on the laws of physics. If an electron has to travel 3 metres it will take 10 "mes longer to reach its des"na"on than one that has to travel 30 cen"metres. This ignores the di&erent layers of so$ware and network hardware that the electron (or request) may have to pass through on its journey. The mainframe supports virtual networking, which is signicantly faster than machine to machine communica"on. Both approaches carry their own benets and disadvantages – and the aim of this paper isn’t to say that scale-up is the only, and best, way to deliver internal cloud. But the idea that scale-out is the only, and best, way is equally untrue. A -  Y: MARSH Marsh is the world’s largest insurance broker and risk advisor, with 24,000 employees worldwide and annual revenues of “ Our cost-benet analysis showed almost $5 billion. Its European, Middle Eastern and African that a distributed architecture based (EMEA) business units have their headquarters in London, on x86 cores would be signicantly and the company employs approximately 11,000 people. more expensive than an IFL engine – which shows that it’s denitely worth The company took the decision to modernize an exis"ng inves!ga!ng the Enterprise Linux mainframe applica"on by developing a modern web-based Server op!on even for rela!vely interface to it. The company chose to use IBM Ra"onal small environments. “ Host Access Transforma"on Services (HATS) to support the Tim Benne! - Project Manager, Marsh web-enablement of the applica"on, and then set about determining the ideal hardware environment for the new func"onality. All contents © The Bathwick Group Ltd 2010 10 www.bathwick.com
  • 11. TheBathwickGroup The mainframe and the cloud - November 2010 The company had ini"ally intended to build a new Linux environment on standard x86-based servers, “ Compared to procuring several new Intel servers, wai!ng for them to be delivered, when IBM proposed using an Enterprise Linux Server nding space for them in the data centre, solu"on instead, using an Integrated Facility for Linux connec!ng all the cables and so on, it was a processor in their System z9 Business Class server, really quick and easy process. We didn’t need they conducted a comparison of both op"ons. The to buy any new hardware, and we completed conclusion of the study was that the mainframe- the whole implementa!on during a standard based solu"on o&ered be!er levels of reliability and weekend maintenance slot, so there was no exibility, at signicantly lower cost to the Intel based disrup!on for our users. ” proposal. Tim Benne! – Project Manager, Marsh Since the company already had an IBM mainframe, the deployment of the new infrastructure was simply a ma!er of upgrading the machine, a task that was completed by an IBM engineer during a scheduled maintenance window. A -  Y:  UY  B The University of Bari (Universita di Bari) in Italy has chosen to build a cloud compu"ng service on top of an IBM System z mainframe running the Linux opera"ng system with DB2, WebSphere and Tivoli middleware from IBM. The goal of the project is to help local businesses, including shermen, winemakers and trucking companies, gain access to compu"ng power that would otherwise be unavailable to them. One example allows local shermen to determine demand in local sh markets in real "me, using a simple touch screen interface installed in their boats. The shermen enter the type of sh caught just minutes ago and instantaneously start a virtual auc"on with wholesalers on the docks. Another project links winemakers at up to 60 co-opera"ve wineries to enable them to determine market demand for various types of wines, then package and ship wines demanding the highest price. The overall project is being undertaken as part of the Daisy-Net consor"um, which combines public and private organisa"ons from ve regions of southern Italy. Daisy-Net researches, develops and provides transfer and training ac"vi"es for new technologies in the informa"on and communica"on technology sector and beyond. Introducing the Enterprise Linux Server – Cloud in a box The Enterprise Linux Server (ELS) is made up of an IBM z-Series mainframe that ships with the hardware and so$ware necessary to create and manage a private cloud of Linux servers. The system takes advantage of the mainframe’s built-in support for virtualisa"on, combined with IBM’s Tivoli Service Automa"on Manager, which provides the management and provisioning framework for the machine. The ELS can also be used to support Linux-based desktop virtualisa"on, using eyeOS, an open-source Virtual Linux Desktop pla#orm. All contents © The Bathwick Group Ltd 2010 11 www.bathwick.com
  • 12. TheBathwickGroup The mainframe and the cloud - November 2010 The key features of the ELS are that it is built on mature technology, and makes extensive use of IBM’s so$ware por#olio (Tivoli, Ra"onal, Systems Director) to provide all of the core infrastructure that you need in order to deploy an “ Allianz, a global insurance company internal cloud – all within a single hardware footprint. based in Germany, e$ec!vely migrated applica!ons from 60 distrib- The system allows users to dene their own standard images uted systems to one Linux on IBM that can drama"cally reduce the "me it takes to deploy System z computer in 48 hours at its new servers. This approach can have a signicant impact on Australian data centre. “ hardware and so$ware costs (IBM has seen savings of up to 75% in some scenarios) as well as reducing the cost and "me associated with systems management. Another by-product of using standardised images is the reduc"on in faults arising from congura"on errors (which in some environments can account for the vast majority of applica"on failures). T E L S    ATOS Atos Origin is a major interna"onal IT services provider, employing 50,000 professionals in 40 countries. The company o&ers integrated design, build and operate solu"ons to large mul"-na"onal clients in targeted industry sectors, and has annual revenues of around £5.8 billion. An important part of Atos Origin’s UK business is managed services – from basic server hos"ng through to full lifecycle management for both hardware and applica"ons. Atos Origin recently developed a new Linux-based applica"on that would help rail transport companies deliver travel informa"on to sta& and passengers via their mobile phones, “ We have been using IFL processors and wanted to deliver it as a shared service for several major for several years now, and in our rail companies. The company also wanted to enhance its opinion, System z has proved itself as service to customers in the insurance industry by upgrading an ideal enterprise Linux pla"orm. “ from MQSeries Integrator (MQSI) middleware to a new IBM Colin Clews - Technologies Manager, Atos WebSphere Message Broker solu"on, which would provide Origin support for Web Services integra"on for a number of core insurance applica"ons. Atos Origin decided to use the IBM Enterprise Linux Server (ELS) to host these systems. The ELS is an IBM System z10 Business Class server that leverages specialised IBM Integrated Facility for Linux (IFL) processors to provide a highly available environment for virtualised Linux workload. “We have been using IFL processors for several years now, and in our opinion, System z has proved itself as an ideal enterprise Linux pla#orm,” says Colin Clews. “With a tradi"onal distributed approach – even using the latest virtualisa"on-enabled x86 processors – you end up with a lot of unreliable boxes to manage. All contents © The Bathwick Group Ltd 2010 12 www.bathwick.com
  • 13. TheBathwickGroup The mainframe and the cloud - November 2010 With the ELS, we can run hundreds of environments within a single physical footprint, and easily deliver the “ The z10 IFL engines in the new ELS not 24x7 availability that our customers demand.” only run Linux environments approxi- mately twice as fast as the previous The addi"onal capacity of the z10 IFLs is available at a genera!on – they also o$er around 40 50 percent lower price than the previous genera"on percent more capacity, enabling us to of System z servers. The result is a price-performance expand our Linux footprint without improvement of up to 65 percent. increasing our costs. “ Running the mobile applica"on under Linux on the ELS Colin Clews - Technologies Manager, Atos proved so successful that Atos also decided to deploy Origin the WebSphere Message Broker solu"on in a similar environment. The company has invested in a second IFL to handle this new workload. “Running WebSphere Message Broker under Linux on the ELS is considerably more cost-e&ec"ve than our exis"ng MQSI and z/OS environment,” says Colin Clews. “As a result, we not only benet from the extended func"onali"es that Message Broker o&ers – we can also deliver them at a more compe""ve price for our customers.” To provide disaster recovery for the System z environment, the produc"on ELS machine is mirrored to a smaller z9 server at a secondary data centre. The z9 server features IBM Capacity Backup Upgrade (CBU), which enables its capacity to be drama"cally increased if and when it is required to take over the produc"on workload from the produc"on server. With CBU, this spare capacity a!racts no maintenance or so$ware licensing costs un"l it is actually ac"vated. Linux on the mainframe is also helping Atos Origin develop an on-demand compu"ng model for its customers, enabling them to pay for compu"ng power on a transac"onal basis. “With z/VM, we can provision new Linux servers very rapidly, and measure capacity usage very precisely,” says Colin Clews. “If customers need more power, we can just turn it on almost instantly, and bill them accordingly. “We think that the on-demand exibility o&ered by Linux on the IBM ELS pla#orm will be very a!rac"ve to customers who currently run large distributed server environments, enabling them to reduce costs and increase responsiveness.” Summary It isn’t the inten"on of this paper to say that the mainframe is the “best” host for all cloud compu"ng environments, but that in some situa"ons we believe that it certainly is. Much of the current discourse around virtualisa"on and cloud compu"ng ignores the power of large scale-up environments like the mainframe and this is simply foolish. The mainframe of today is profoundly di&erent from the mainframe of the 1980s and when deba"ng the ideal pla#orm for cloud environments, it is only fair to compare today’s mainframe with non-mainframe technology rather than relying on a dated understanding of the mainframe’s value proposi"on. All contents © The Bathwick Group Ltd 2010 13 www.bathwick.com
  • 14. TheBathwickGroup The mainframe and the cloud - November 2010 About The Bathwick Group The Bathwick Group is a research-based consul"ng company that helps clients address their most pressing needs in strategic planning, go-to-market planning and execu"on, and IT infrastructure e&ec"veness: Consul!ng " Bathwick Engage A rapid collabora"ve consul"ng service combining external experts and IP protec"on mechanisms to expedite solu"ons to major corporate challenges " Enterprise IT strategy Planning and contract support for enterprise IT leaders; produc"vity and infrastructure agility benchmarking and best prac"ce " IT vendor strategy and marke!ng Customer analysis and deep research for IT vendors targe"ng mid-market and enterprise markets " Sustainability strategy Modelling and benchmarking for organisa"ons wishing to embed sustainable prac"ces and mi"gate strategic risks Research and benchmarking " Research Pla"orm A exible so$ware pla#orm for organisa"ons to deploy their own surveys and benchmarks, both internal (e.g. employee surveys), or external (e.g. market intelligence) " Media Pla"orm A research pla#orm for online media customers, providing an engagement environment to encourage reader registra"on and generate new revenues Sales enablement " The Customer Insight Pla"orm A so$ware pla#orm that supports deployment of sales guidance, customer assessments, marke"ng collaterals, ROI tools, etc. direct to a seller’s laptop for online and o'ine opera"on; for both direct and channel sales support. The Bathwick Group also includes the ThinkAgain Partnership LLP, a global collabora"ve research network, which brings together academics, writers, business and poli"cal leaders to generate new insights into business produc"vity and perform- ance, geo-poli"cal and environmental issues. LVL03001-USEN-00 All contents © The Bathwick Group Ltd 2010 14 www.bathwick.com