Today HP announced a new set of technology programs and future products designed to move x86 server technology for both Windows and Linux more fully into the realm of truly mission-critical computing. My interpretation of these moves is that it is both a combined defensive and pro-active offensive action on HP’s part that will both protect them as their Itanium/HP-UX portfolio slowly declines as well as offer attractive and potentially unique options for both current and future customers who want to deploy increasingly critical services on x86 platforms.
Bearing in mind that the earliest of these elements will not be in place until approximately mid-2012, the key elements that HP is currently disclosing are:
ServiceGuard for Linux – This is a big win for Linux users on HP, and removes a major operational and architectural hurdle for HP-UX migrations. ServiceGuard is a highly regarded clustering and HA facility on HP-UX, and includes many features for local and geographically distributed HA. The lack of ServiceGuard is often cited as a risk in HP-UX migrations. The availability of ServiceGuard by mid-2012 will remove yet another barrier to smooth migration from HP-UX to Linux, and will help make sure that HP retains the business as it migrates from HP-UX.
Analysis engine for x86 – Analysis engine is internal software that provides system diagnostics, predictive failure analysis and self-repair on HP-UX systems. With an uncommitted delivery date, HP will port this to selected x86 servers. My guess is that since the analysis engine probably requires some level of hardware assist, the analysis engine will be paired with the next item on the list…
Since cloud became a household word, vendors and enterprises alike have jumped to declare victory on cloud with services and infrastructure implementations that really don't deliver cloud value but have the same foundation - something we call "cloudwashing." This is a dangerous gambit as you claim legitimacy but don't activate the same economics, deliver the autonomy that cloud services offer to your internal users and aren't standardized or automated enough to deliver transformative agility. In other words you claim cloud but are achieving only incrementally better value.
This week AMD finally released their AMD 6200 and 4200 series CPUs. These are the long-awaited server-oriented Interlagos and Valencia CPUs, based on their new “Bulldozer” core, offering up to 16 x86 cores in a single socket. The announcement was targeted at (drum roll, one guess per customer only) … “The Cloud.” AMD appears to be positioning its new architectures as the platform of choice for cloud-oriented workloads, focusing on highly threaded throughput oriented benchmarks that take full advantage of its high core count and unique floating point architecture, along with what look like excellent throughput per Watt metrics.
At the same time it is pushing the now seemingly mandatory “cloud” message, AMD is not ignoring the meat-and-potatoes enterprise workloads that have been the mainstay of server CPUs sales –virtualization, database, and HPC, where the combination of many cores, excellent memory bandwidth and large memory configurations should yield excellent results. In its competitive comparisons, AMD targets Intel’s 5640 CPU, which it claims represents Intel’s most widely used Xeon CPU, and shows very favorable comparisons in regards to performance, price and power consumption. Among the features that AMD cites as contributing to these results are:
Advanced power and thermal management, including the ability to power off inactive cores contributing to an idle power of less than 4.4W per core. Interlagos offers a unique capability called TDP, which allows I&O groups to set the total power threshold of the CPU in 1W increments to allow fine-grained tailoring of power in the server racks.
Turbo CORE, which allows boosting the clock speed of cores by up to 1 GHz for half the cores or 500 MHz for all the cores, depending on workload.
Forrester has done quite a number of reports in the last two years around platform-as-a-service (PaaS) from the long-term strategy perspective from me and from the application developer perspective from my friend John R. Rymer. During this time, we saw many different business cases around PaaS. We have predicted and quantified that the major buying power of PaaS will come out of three camps:
ISVs are buying PaaS technology. This is a model that we saw with many ISVs on major platforms that managed to create a viable marketplace such as salesforce.com's AppExchange and Google's marketplace.
Corporate application developers are using PaaS to deploy custom apps and add-ons around SaaS applications. They are doing this significantly faster and at a lower TCO than before.
Forrester just published parts I & II of its market overview of the public cloud market and these reports, written primarily for the Infrastructure & Operations (I&O) professionals, reveal as much about you – the customers of the clouds – as it does about the clouds themselves.
As discussed during our client teleconference about these reports, clearly the Infrastructure as a Service (IaaS) market is maturing and evolving and the vendors are adapting their solutions to deliver greater value to their current customers and appeal to a broader set of buyers. In the case of pure clouds such as Amazon Web Services, GoGrid and Joyent, the current customers are developers who are mostly building new applications on these platforms. Their demands focus on enabling greater innovation, performance, scale, autonomy and productivity. To broaden the appeal of their cloud services, they aim to deliver better transparency, monitoring, security and support – all things that appeal more to I&O and security & risk managers (SRM).
Emerging ARM server Calxeda has been hinting for some time that they had a significant partnership announcement in the works, and while we didn’t necessarily not believe them, we hear a lot of claims from startups telling us to “stay tuned” for something big. Sometimes they pan out, sometimes they simply go away. But this morning Calxeda surpassed our expectations by unveiling just one major systems partner – but it just happens to be Hewlett Packard, which dominates the WW market for x86 servers.
At its core (unintended but not bad pun), the HP Hyperscale business unit Project Moonshot and Calxeda’s server technology are about improving the efficiency of web and cloud workloads, and promises improvements in excess of 90% in power efficiency and similar improvements in physical density compared with current x86 solutions. As I noted in my first post on ARM servers and other documents, even if these estimates turn out to be exaggerated, there is still a generous window within which to do much, much, better than current technologies. And workloads (such as memcache, Hadoop, static web servers) will be selected for their fit to this new platform, so the workloads that run on these new platforms will potentially come close to the cases quoted by HP and Calxeda.
You know there are developers in your company using public cloud platforms, but do you really know what they are doing? You suspect it’s just test and development work, but are you sure? And if it is production workloads are they taking the steps necessary to protect the company? We have the answers to these questions and you may be surprised by how far they are going.
It’s tough being an infrastructure & operations professional these days. According to our ForrSight surveys, for every cloud project you know about there could be 3 to 6 others you don’t know about. Business unit leaders, marketing and sales professionals and Empowered developers are leading the charge. They aren’t circumventing I&O as a sign of rebellion – they simply are trying to move quickly to drive revenue and increase productivity. While every I&O professional should be concerned about this pattern of shadow IT and its implications on the role of I&O in the future, the more immediate concern is about whether these shadow efforts are putting the company at risk.
The bottom line: Cloud use isn’t just test and development. In fact, according to our ForrSight research there’s more production use of IaaS cloud platforms than test and development and broader use is coming (see Figure 1 below). The prominent uses are for training, product demonstration and other marketing purposes. Our research also shows that test and development projects in the cloud are just as likely to go to production in the cloud as they are to come back to your data center.
It’s easy to bash incumbent telcos, to count them as being among the losers in the digital revolution. Cloud services players are taking business from telcos in the storage and server capacity space. Over-the-top providers are free-riding on the telco infrastructure. Software firms are eating into the communication business. Regulators are pressing for further price reductions. And to top this scenario, telcos are continuing to undercut each other in price wars.
During a round of executive discussions with Forrester, Orange Business Services (OBS) has shown that against these odds, it keeps a pretty even keel regarding the most hyped topics in ICT, most notably cloud and mobility. OBS is selective in its cloud offerings, focusing on UCaaS and IaaS. UCaaS is a natural extension of its communication business and thus falls into OBS’ home turf. All telcos should see communication services from the cloud as a natural extension of what they have always done.
OBS’ drive into IaaS, meanwhile, looks like a less convincing pitch. Its IaaS offering essentially comprises a virtual data centre offering with virtual firewalls and load balancing. The question is: How OBS can compete against the dominant cloud players in the storage and server space? In the short term, such an approach is conceivable. However, OBS will need to provide a much broader range of virtual infrastructure choices to avoid slipping into a low-margin market segment.
In a surprising move, HP and Cisco announced that HP will be reselling a custom-developed Cisco Nexus switch, the “Cisco Nexus B22 Fabric Extender for HP,” commonly called a FEX in Cisco speak. What is surprising about this is that the FEX is a key component of Cisco’s Nexus switch technology as well as an integral component of Cisco’s UCS server product, the introduction of which has pitted the two companies in direct and bitter competition in the heart of HP’s previously sacrosanct server segment. Combined with HP’s increasing focus on networking, the companies have not been the best of buds for the past couple of years. Accordingly, this announcement really makes us sit up and take notice.
So what drove this seeming rapprochement? The coined word “coopetition” lacks the flavor of the German “Realpolitik,” but the essence is the same – both sides profit from accommodating a real demand from customers for Cisco network technology in HP BladeSystem servers. And like the best of deals, both sides walk away thinking that they got the best of the other. HP answers the demands of what is probably a sizable fraction of their customer base for better interoperability with Cisco Nexus-based networks, and in doing so expects to head off customer defections to Cisco UCS servers. Cisco gets both money (the B22 starts at around $10,000 per module and most HP BladeSystem customers who use it will probably buy at least two per enclosure, so making a rough guess at OEM pricing, Cisco is going to make as much as $8,000 to $10,000 per chassis from HP BladeSystems that use the B22) from the sale of the Cisco-branded modules as well as exposure of Cisco technology to HP customers, with the hope that they will consider UCS for future requirements.
OK, out of respect for your time, now that I’ve caught you with a title that promises some drama I’ll cut to the chase and tell you that I definitely lean toward the former. Having spent a couple of days here at Oracle Open World poking around the various flavors of Engineered Systems, including the established Exadata and Exalogic along with the new SPARC Super Cluster (all of a week old) and the newly announced Exalytic system for big data analytics, I am pretty convinced that they represent an intelligent and modular set of optimized platforms for specific workloads. In addition to being modular, they give me the strong impression of a “composable” architecture – the various elements of processing nodes, Oracle storage nodes, ZFS file nodes and other components can clearly be recombined over time as customer requirements dictate, either as standard products or as custom configurations.