Organisations tend to have one of two IT strategies today: those who are already planning and eventually implementing cloud strategy, and those who are going to be doing it soon. But, the options that companies are faced with are dizzying, often contradictory, and usually dangerously expensive. So what’s the best way for organisations to find the ideal cloud service for their specific needs?
Determining what is needed from the cloud will drive what platform organisations should deploy on. Considerations like budget, expected performance, and project timeline all have to be carefully balanced before plunging ahead. Broadly speaking the platform options range from using someone else’s public cloud, such as AWS, to building your own private cloud from scratch. Where an organisation lands on that spectrum will be driven by how they rank the primary factors involved.
In this guest blogpost, Christopher Aedo, chief product architect at Mirantis explains how to evaluate the cloud requirements and pick the right platform
In essence there are seven key factors to address that will help businesses clarify what really matters and enable them to establish their individual cloud requirements. These are:
Control: How much control do you have over the environment and hardware? Make sure the cloud platform you select delivers the level of control you require.
Deployment Time: How long before you need to be up and running? How much time will you burn just sorting out, ordering, racking and provisioning the hardware? It is critical that the cloud platform you choose can deployed in the right amount of time.
Available Expertise : Can your single IT staff member handle the project, or do you need a team of experts poached from the biggest cloud makers? Choose a cloud platform that matches the expertise you have available – or you can afford to bring in.
Performance: In a single server there are so many components impacting performance – from the memory bus to your NIC and everything in between. However performance directly correlates with budget – a larger budget will usually see greater performance. However there is no reason a smaller budget can’t see high performance – providing you select the right option.
Scalability: Your platform of choice should accommodate adding, or reducing, capacity quickly and easily. Will your chosen platform require downtime to scale up or down or can it be executed seamlessly?
Commitment: From no contract “utility pricing” to the long term investment of owning all your gear – the longer you’re tied up, the greater the risk.
Cost: This may be the most important and most difficult factor to account for. You can see it as an output from your other factors, or your ultimate limiter dictating where you’ll make concessions. There are definitely some good ways to maximize your dollar while minimisng your risk as long as you keep your head up and your eyes open.
By addressing these factors early on in the process of implementing a cloud based solution you will save yourself time, resource and budget in the long run. However having addressed what you want the cloud to deliver it is important that you match your requirements with the right type of cloud platform.
Here are the main cloud options:
Option 1: The Public Cloud
The big players here are AWS and RackSpace, but there are other contenders with fewer bells and whistles like DigitalOcean and Linode. These represent the lowest entry barrier (you just need ‘net access and a credit card!) but also offer the least control and the greatest cost increases as you scale up.
The public cloud is priced like a utility offering the opportunity to scale up/down as needed. This is well suited to handling a highly elastic demand, but it’s important to keep an eye on what you’ve spun up.
With a public cloud you get limited access to the underlying hardware, and no visibility into what’s beneath the covers of the cloud – although you will get some flexibility in configuration and near instant deployment of service without the need for any real expert to be involved.
However, generally speaking, you’re going to find relatively low performance with a public cloud with higher performance coming at significantly increased cost. You can also expect to be billed by the minute in return for not being held to any contract. Many will offer discounts with a commitment of some sort, but then you give up the ability to drop resources when you no longer need them.
Option 2: Hosted Private Cloud
There are many well-known vendors offering options in this space, ranging from complete turn-key environments to build-to-order approaches. They will provide the hardware on short-term lease, and will charge you to manage that hardware.
Companies like RackSpace will work with you to architect your environment, and provide assistance in deployment – which could take up to 6 weeks. You’ll need moderate to extreme expertise and your average junior sys-admin is going to be way out of their depth using s uch a service.
Levels of control will vary from high to minimal depending on how much of the platform you manage and deploy yourself. The level of commitment will also vary but the longer you commitment the more likely an alternative platform is to make sense. HPC is not well suited to an elastic demand and upscale in 2-6 weeks – and generally there will be no ‘scale-down’ option.
Option 3: Build your own private cloud (BYPC)
BYPC requires a high level of technical expertise within the business and will present you with the greatest technical and financial risk. However, you will have total control over the hardware design, the network design, and how your cloud components are configured – but expect this to take a year to 18 months to complete.
Your costs in the build-your-own approach can be kept down if performance and reliability are of no concern, or they can (needlessly) go through the roof if you’re not making carefully planned decisions. The performance of BYPC will be entirely dependent on your budget constraints and how successful your architectural planning is.
There are lots of moving pieces, and the risks are tremendous, as you may be committing hundreds of thousands of dollars to your cloud pilot. Ask anyone who’s actually tried this; it’s a lot harder than it looks.
Option 4: Private-cloud-as-a-Service (PCaaS)
PCaaS, such as OpenStack, represents a balance between the value and flexibility of public cloud and the control of private cloud.
PCaaS provides total control over how hardware is used, and that hardware is 100% dedicated to you with a minimum One-day commitment on a rolling contract. As a result of the minimal commitments it can be deployed within a few hours and you will be free to scale the size of your environment up and down at nearly the same pace as if you were on a public cloud.
The costs are higher than a comparable number of VMs in a public cloud, but with no long-term commitment and clear pricing from the start, your financial risks are lower than any other private cloud approach.
You’ll need a moderate skill level with PCaaS but your risks are mitigated because you’re in a managed environment. Whereas, until recently, PCaaS required you to have a reasonable amount of OpenStack knowledge developments such as OpenStack Express have drastically reduced the expertise needed to implement a PCaaS.
Each of these cloud platforms has validity, as well as a real sweet spot, where that particular approach is the only obvious good choice for your business needs. If you properly consider your requirements and how they match with the options available, your cloud project will not end up as a costly mistake.
Microsoft Azure cloud service status website at 5pm BST on Friday, September 26th showed that while the core Azure platform components were working properly, there was “partial performance degradation” on Azure’s HDInsight service for customers in West Europe.
The status website warned that customers may experience intermittent failures when attempting to provision new HDInsight clusters in West Europe.
HDInsight is a Hadoop distribution powered by the cloud. It allows IT to process unstructured or semi-structured data from web clickstreams, social media, server logs, devices and sensors to analyse data for business insights.
Microsoft has assured cloud users that the engineers have identified the root cause of the performance degradation issue and are working out the mitigation steps.
The company has vowed to provide updates every two hours or as events update. I sense a long wait before the weekend beckons for European enterprises’ Azure users.
Doesn’t the NHS use Microsoft Azure HDInsight? Oh yes, it does!
Two large, very large companies that have been under tremendous pressure in the software-defined storage and cloud era – EMC and HP – toyed with the idea of a merger, according to Wall Street Journal but eventually the idea fell apart because of concerns from both HP and EMC on whether their shareholders would give it a nod.
The deal would have created a mega-vendor worth $130bn with HP’s Meg Whitman as the chief executive of the combined entity and EMC’s Joe Tucci as chairman or president.
Ailing EMC has been under pressure from its investors calling for it spin-off its sister company VMware on the prospect that the company will do better if split up.
According to WSJ, the EMC-HP merger talks have been going on for almost a year.
But the combination of two traditional vendors would have only meant more of the same old legacy, complex, slow and big IT offerings. There is an absence of a meaningful synergy but lot of service overlaps.
HP has a bad history in acquisitions. A merger would have been bad news for both companies even though EMC has a better track record of acquisitions and is attempting to redefine itself in the new cloud era.
EMC Corp is far more than EMC – it has all those fingers in the pies of VMware, RSA, VCE, Pivotal and so on: unpicking these or keeping them going forward would be difficult.
Other names mentioned in a merger with EMC include Dell and Cisco Systems.
Mergers are always a hit or miss and more of a risk when the stakes are higher, as in this case. The problem with these traditional vendors is that, in the past, they have tried to address all the aspects of a datacentre and so they have competing products. For example, EMC-owned VMware’s software-defined network (SDN) offerings threaten Cisco’s switches and routers business worth billions.
As one analyst tells me if EMC is really seeking a merger, it should be going for a Rackspace-type platform company (not Rackspace itself as it has ruled itself out now) where EMC can make a bigger play of VMware’s cloud offering, of the whole software-denied everything message, of ViPR and so on.
Or would Tucci go for Cisco? Markets are betting on EMC-Cisco deal with EMC share prices are up 16 cents.
A merger with heavyweight HP would have left a company trying to sell a complex approach into standard customers’ datacentres. Thankfully, it was only a thought.
I had a chance to see Michael Dell – in flesh – for the first time yesterday in Brussels at the Dell Solutions Summit. He delivered a great keynote on Dell’s datacentre strategy, its investment plans and also spoke about all the hot IT topics – software-defined infrastructures, internet of things, security and data protection.
Michael sounded optimistic about Dell’s place in the future IT but what was new was how open Dell has become as a company and its firm commitment to all things that determine the new age IT – software defined, cloud, security, mobile, big data, next-gen storage and IoT.
For one, Michael was candid with the numbers. He said:
- The total IT market is worth $3 trillion and we have a 2% share of it. Only 10 companies have 1% or more share of that $3 trillion market.
- Dell’s business comprises 85% government and enterprise IT and just 15% is end-user focused.
This kind of number-feeding the press and analysts is new at Dell, which, until now, like rest of the service providers in the industry kept business numbers close to its chest.
But that was not all, Michael didn’t hold back from saying a few things that raised a few eyebrows:
- “I wish we hadn’t made some of the acquisitions we did.”
- “As ARM moves to 64-bit architecture, it becomes more interesting,” Michael said. He said the company is open to working with its longstanding partner Intel’s rival for mainstream datacentre products if that’s where the market moved.
- He also said, Dell is a big believer of the software-defined future. “We ourselves are moving our storage IT into a software defined environment.”
- And to those that wrote off the PC industry, Michael said: “We absolutely believe in the PC business, we are consolidating/growing”.
Michael’s optimism and confidence in the company’s future is a far cry from last year when the company’s ailing business strategy forced it to get itself off the public eye.
“Going private has helped us,” he said while speaking in Brussels. “It has enabled us to put our focus 100% on our customers. We have invested more in research, development, innovation and in channels in the past year.”
Dell also seems to be striking a right chord with its customers, channel and analysts as those I spoke to said they like the company a lot and are pleased with how quickly it adapts and listens to its users.
Dell Research will be focused on five areas – software defined IT, next generation storage (NVM, Flash), next gen cooling, big data/analytics, IoT. Analysts say that’s a good bet.
“Dell’s foray into research clearly designed to establish it as an IT innovator as well as a scale/efficiency player,” says Simon Robinson from 451 Research group on Twitter.
Product-wise too it is making progress. Dell has been more creative than its competitors in designing its new servers on the latest Xeon chip. Its 13th-generation PowerEdge servers have capabilities such as NFC for server inventory management, new Flash capabilities, and has more front sockets.
Dell is also being innovative in its enterprise cloud strategies. It is providing the reference architectures, proof of concepts and server technologies to its system integrators to do the cloud implementation for customers. Having catered to the likes of AWS in the past, Dell has used that cloud experience to build reference architectures but gets the channel to implement it.
“We see private cloud as the future of cloud computing,” Michael said. According to him, enterprises in Europe prefer “local” clouds for data sovereignty and privacy issues, so it is supporting local system integrators with local datacentres to build cloud for the customers.
Michael and his company are certainly making the right noises and are investing in the right technologies. But will it increase their ranking in the datacentre (which I see as fourth after Cisco, HP and IBM – in that order), only time will tell.
Also, is it symbolic that Dell held its Solutions Summit party at the world-famous Comic Strip Museum in Brussels – the home of Tintin, Captain Haddock, the Smurfs and Asterix? Don’t know, but I sure did have fun!
Ambitious startups and developers around the world got a big treat from Google ahead of the weekend – $100,000 worth of Google cloud credits along with 24/7 support from the tech experts at Google.
Urs Hölzle, Google Fellow launched the “Google Cloud Platform for Startups” initiative on Friday to help startups take advantage of its enterprise cloud offering and “get resources to quickly launch and scale their idea“. The free cloud resources are aimed at helping developers focus on code without worrying about managing infrastructures. Google has also not set any restrictions on the type of cloud services users can spend their credits on giving them complete flexibility to choose IaaS, PaaS or SaaS or even data-related cloud offerings.
But to qualify, the startups will need to have less than $5m dollars in funding and have less than $500,000 in annual revenue. And the cloud credits are available through incubators, accelerators and investors.
Cloud computing has always been a technology that democratized IT by giving startups a level-playing field to compete with the big players. And cloud behemoth AWS is seen as the “go-to” cloud option for the cool, emerging poster-children of the web such as Netflix and Instagram (before it was acquired by Facebook).
Google Appliance as shown at RSA Expo 2008 in San Francisco. It was only a computer case with no parts inside.-Daniel A (Photo credit: Wikipedia)
Google, AWS, Microsoft and IBM have so far been tripping over each other to announce price drops to lure more users to their cloud services. AWS launched a free programme called AWS Activate to help selected startups with resources for working with AWS. It includes services such as AWS web-based training, virtual office hours with an AWS Solutions Architect and credit for eight self-paced training labs.
But Google has now upped the game by targeting startups with cloud credits of the size and scale that hasn’t been seen before.
Cloud giants are targeting the ambitious startups because today’s startups can become tomorrow’s enterprises and the providers want these potential customers to use their platforms.
Startups are lean and are quick at adopting new technologies such as the cloud but need some technical expertise so they can focus on their business than the underlying technology. Google is offering exactly that in a bid to get a footprint in the enterprise IT.
It will be interesting to see how UK’s promising tech startups such as Swiftkey and Hailo use these cloud credits offered by Google. But more importantly how quickly AWS, Microsoft and IBM respond to this.
Cloud wars just got spicier!
Cloud computing is becoming a default option of delivering IT services but to reap all the benefits of the cloud, enterprises must do the boring stuff first.
On Thursday, I attended a Westminster eForum seminar on the future of cloud computing where I witnessed very interesting conversations around cloud adoption, risks, and its future from speakers ranging from analysts, legal experts and industry association heads to cloud vendors and public sector professionals.
When experts said cloud can be secure and cost-effective and can lead to innovation – it did not raise any eyebrows from the delegates. This suggests to me that users are fully convinced of cloud’s benefits.
But even then, some cloud projects backfire. Why?
The excitement of cloud is leading enterprises to overlook the boring work they need to do beforehand to yield the full benefits of the cloud. Ovum analyst Gary Barnett illustrated this best in his (PowerPoint-free!) session. Here is an article where Gary shares the user instances where cloud has failed.
“My mum made sure I ate my broccoli before I got my pudding,” Gary said. But in the cloud world, no one’s eating the broccoli, he said.
“If you don’t clean up your data before putting it on the cloud platform, you will have cloudy rubbish.” He also pointed that some users are finding cloud expensive because they are not building proper policies and guidelines around its use.
Experts at the seminar insisted cloud is a secure way of doing IT and cloud breaches are usually because of users’ “silly and predictable passwords” and their lack of awareness. Gary urged enterprises to educate users on the loopholes of predictable passwords.
“No one loves the boring stuff. But just like you have to eat your greens, you have to do all the boring stuff before adopting the cloud. Otherwise you’re just transferring onsite mess offsite,” Gary said.
The “Eat your greens” theme continued throughout the seminar and the floor roared out laughing when Microsoft’s cloud director Maurice Martin said: “In my case, the greens were the cabbages, broccoli was too posh.”
One of the main barriers to cloud adoption is data privacy. This is an issue because, for the majority of cloud providers, EU/EEA and US data privacy and Information Security standards are minefields which are very difficult to cross. And that is because their focus has been on the ease of use and functionality of their services rather than the all-important data privacy, information security, data integrity and reliability requirements around providing these services responsibly.
But, when looking through the plethora of cloud service providers, you can immediately sort the ‘wheat from the chaff’ once you start drilling down into the data privacy, information security, data integrity and reliability capabilities offered to ensure the protection of your and your customers’ data.
In this guest blog post, Mike McAlpen, the executive director of security & compliance and data privacy officer at 8×8 Solutions outlines the questions cloud users must ask their providers before signing a contract.
Have you chosen the right cloud services provider?
– Mike McAlpen
By asking your cloud services provider the following questions you will be on the way to knowing whether you can entrust your data into its care.
- Compliance with EU/EEA data privacy standards
The most important question is whether your provider can provide third-party verification/audit assurance of their compliance with EU/EEA and/or US data privacy standards? It is not enough for the provider to simply produce this verification/audit assurance, it must show that it has fully implemented the UK Top 20 Critical Security Controls For Cyber Defence and/or ISO 27001 and/or rigorous US standards such as the Federal Information Security Act (FISMA) and International Information PCI-DSS v3.0 security standards.
If this verification/audit assurance is not available then your business is at peril of not meeting EU/EEA and/or US standards.
In the US, many EU/EEA and other countries it can be a criminal offence if a breach of personal data privacy occurs and an individual employee or senior management, depending on the circumstances of the breach, is deemed to be responsible.
- Onward Transfer of Data
Does your provider work with third-party suppliers in order to deliver the cloud services it offers? If so you must check that it has contracts in place with its third-party suppliers that provide assurance that they are, and will continue to be, compliant with EU/EEA and/or US standards.
- Data Encryption
Does the cloud solutions vendor provide the capability to encrypt sensitive data when it is being transferred across the Internet and importantly again when it is ‘at rest’? (i.e. Stored by your cloud services provider, or in files on a computer, laptop USB flash drives or other electronic media?).
- The Right to be Forgotten
Has your provider’s solution been engineered to enable it to identify and associate each user’s personal information data? It must also provide the capability for each user to view and modify this personal data. In addition, if the user wishes this data to be deleted, the provider must then be able to completely erase all of that person’s personal data without affecting anyone else’s data.
- Service Level Agreements (SLAs)
Outside of compliance with data privacy standards, another key issue is asking your provider how you will determine and then document, within your services contract, the required service level agreements (SLAs). It’s no use whatsoever having the cloud services you have always wanted if you have no way of measuring or monitoring if they are actually being delivered to an acceptable level or if there are no financial penalties for non-compliance.
If your provider cannot answer “yes” to the above questions and you cannot agree to mutually acceptable SLAs – look for another provider!
On Day 1 of its annual conference VMworld 2014 themed “No Limits”, VMware unveiled its strategies around open cloud platform OpenStack and around container technology Kubernetes. It also launched new tools to extend its software-defined datacentre and hybrid cloud offerings.
Open software-defined datacentre
One of the significant announcements was the VMware Integrated OpenStack – a service that provides enterprises – especially SMBs the flexibility to build a software-defined datacentre on any technology platform (VMware or not).
VMware Integrated OpenStack distribution is aimed at helping customers repatriate workloads from “unmanageable and insecure public clouds”. Take that AWS.
Container technology and VMware infrastructures; Kubernetes collaboration
VMware is collaborating with Docker, Google and Pivotal to allow enterprises to run and manage container-based applications on its platforms.
At the annual conference, VMware said it has joined the Kubernetes community and will make Kubernetes’ patterns, APIs and tools available to enterprises. Kubernetes, currently in pre-production beta, is an open-source implementation of container cluster management.
With Google, VMware’s efforts will focus on bringing the pod based networking model of Open vSwitch to enable multi-cloud integration of Kubernetes.
Not only will deep integration with the VMware product line bring the benefits of Kubernetes to enterprise customers, but their commitment to invest in the core open source platform will benefit users running containers,” said Joerg Heilig, VP Engineering, Google Cloud Platform. “Together, our work will bring VMware and Google Cloud Platform closer together as container based technologies become mainstream.”
With Docker, it will collaborate to allow Docker Engine on VMware workflows. It will also work to improve interoperability between Docker Hub with VMware vCloud Air, VMware vCenter Server and VMware vCloud Automation Center.
New hybrid cloud capabilities
At VMworld, VMware released new hybrid cloud service capabilities and a new line-up of third-party mobile application services. The new capabilities include vCloud Air Virtual Private Cloud OnDemand that offers customers with on demand access to vCloud Air. Another capability – VMware vCloud Air Object Storage – is aimed at providing users with scalable storage options for unstructured data. It will enable customers to easily scale to petabytes and only pay for what they use, according to the company.
It also launched mobile development services within VMware’s vCloud Air’s service catalog.
Management as a service offerings
VMware also released two new IT management tools under its vRealize brand- for managing a software-defined datacentre and public cloud infrastructure services (IaaS).
VMware vRealize Air Automation is the cloud management tool that allows users to automate the delivery of application and infrastructure services while maintaining compliance with IT policies.
Meanwhile, VMware vRealize Operations Insight offers performance management, capacity optimization, and real-time log analytics. The tool also extends operations management beyond vSphere to an enterprise’s entire IT infrastructure. Another sign than VMware is opening up its ecosystem to accommodate other virtualisation platforms.
Partnerships with Dell on software defined services
VMware has extended collaboration with Dell to combine its NSX network virtualisation platform with the latter’s converged infrastructure products.
“Global organisations are adopting the software-defined datacentre as an open, agile, secure and efficient architecture to simplify IT and transition to the hybrid cloud,” said Raghu Raghuram, executive vice president, SDDC division, VMware. “The software-defined datacentre enables open innovation at speeds that cannot be matched in the hardware-defined world. As partners, VMware and Dell will advance networking in the SDDC, and collaborate to make advanced network virtualisation available to mutual customers.”
Partnership with HP on hybrid cloud
VMware and HP have extended their collaboration to give momentum to users’ SDDC and hybrid cloud adoption. As part of the partnership, HP Helion OpenStack will support enterprise-class VMware virtualisation technologies.
The companies will also make standalone HP-VMware networking solution generally available. Together, these collaborative efforts can help simplify the adoption of the software-defined datacentre and hybrid cloud with less risk, and with greater operational efficiency and lower costs.
All in all, looks like VMware is opening up to competitive platforms and warming up to open source technologies but retains its standoffish traits when it comes to public cloud services.
Just when I thought to myself: Cloud services must be improving as there are fewer outages reported this year than there were last year, Microsoft Azure cloud service went down for many users, including European ones, earlier this week.
Microsoft’s Azure status page currently displays a chirpy:
Everything is running great.
It also displays a bright green check besides its core Azure platform components such as Active Directory, and popular cloud services including its SQL Databases, and storage services.
A snoop into its history page shows that all wasn’t good aboard Azure on Monday and Tuesday. Users experiencing full service interruption and performance degradation across several services including StorSimple, storage services, website services, backup and recovery and virtual machine offerings.
For a brief moment on Tuesday, August 19th, a subset of its customers in West Europe and North Europe using Virtual Machines, SQL Database, Cloud Services, and Storage were unable to access Azure resources or perform management operations. Users accessing Azure’s Website cloud services in Northern Europe too faced connectivity issues.
WELCOME TO Microsoft® (Photo credit: Wikipedia)
The previous day, some of its customers across multiple regions were unable to connect to Azure Services such as Cloud Services, Virtual Machines, Websites, Automation, Service Bus, Backup, Site Recovery, HDInsight, Mobile Services, and StorSimple.
Some of the services were down for almost five hours.
This week’s global outage follows last week’s (August 14th) Azure outage where users across multiple regions experienced full service interruption to its Visual Studio Online. The news doesn’t bode well for CEO Satya Nadella’s “cloud-first” strategy.
Well, I may have tempted fate. Resilience and reliability are two words I’ll use sparingly to describe public cloud services.
Internet of Things, big data, and social media are all creating an insatiable demand for scalable, sophisticated and agile IT resources, making datacentres a true utility. This is making big tech and telecom companies to drift a bit from their core competency and build their own customised datacentres – take Telefonica’s €420m investment in its new Madrid datacentre.
But the mind-boggling growth of computing infrastructure is occurring amid shocking increases in energy prices. Datacentres consume up to 3% of global electricity and produce 200 million metric tons of carbon dioxide, at an annual cost of $60bn. No wonder, IT energy efficiency is primary concern for everyone from CFOs to climate scientists.
In this guest blog post, Dave Wagner, TeamQuest’s director of market development with 30 years of experience in the capacity management space explains why enterprises must not be too hung up on PUE alone to measure their datacentre efficiency.
Measuring datacentre productivity? Go beyond PUE
-by Dave Wagner
In their relentless pursuit of cost effectiveness, companies measure datacentre efficiency with power utilization effectiveness (PUE). The metric measures the total amount of power coming onto the datacentre floor, divided by how much of that power is actually used by the computing equipment.
PUE = Total energy
PUE is necessary but not a sufficient indicator to gauge the costs associated with running or leasing datacentres.
While PUE is a detailed measure of datacentre electrical efficiency, it is one of several elements that actually determine total efficiency. In the bigger picture, focus should be on more holistic and accurate measures of business productivity, not solely on efficient use of electricity.
Gartner analyst Cameron Haight talked about how a very large technology company owns the most efficient datacentre in the world with a PUE of 1.06. This basically means that 94% of every watt that comes into the floor actually gets to processing equipment. This remarkably efficient PUE achievement does not detail what they do with all of that power, and how much total work is accomplished. If all that power is going to servers that are switched on but essentially idling and not actually accomplishing any useful work, what does PUE really tell us? Actual efficiency in terms of doing real-world work could be nearly zero even when the PUE metric indicates a well-run datacentre in isolation.
Datacenter (Photo credit: Wikipedia)
Boiled down, what companies end up measuring with PUE is how efficiently they are moving electricity around within the datacentre.
By some estimates, many datacentres are actually only using 10-15% of their electricity to power servers that are actually computing something. Companies should minimize costs and energy use, but nobody invests in a company solely based on how efficiently they move electricity.
Datacentres are built and maintained for their computing capacity, and for the business work that can be done thereupon. I recommend correlating computing and power efficiency metrics with the amount of useful work and with customer or end user satisfaction metrics. When these factors are optimised in a continuous fashion, true optimization can be realised.
I’ve talked about addressing power and thermal challenges in datacentres for over a decade, and have seen progress made – recent statistics show a promising slowdown in datacentre power consumption rates in the US and Europe due to successful efficiency initiatives. Significant improvements in datacentre integration have helped IT managers control the different variables of a computing system, maximising efficiency and preventing over- or under-provisioning, both having obvious negative consequences.
An integrated approach to planning and managing datacentres enables IT to automate and optimise performance, power, and component management with the goal of efficiently balancing workloads, response times, and resource utilisation with business changes. Just as the IT side analyses the relationships between the components of the stack–networking, server, compute, and applications–the business side of the equation must always be an integral part of these analyses. Companies should always ask how much work they are accomplishing with the IT resources they have; unfortunately, often easier said than done. In the majority of datacentres and connected enterprises, the promise of continuous optimisation has not been fully realised, leaving lots of room for improvement.
As datacentres grow in size and capabilities, so must the tools used to manage them. Advanced analytics have become essential to bridging IT and business demands, starting with relatively simple co-relative and descriptive methods and progressing through predictive to prescriptive approaches. Predictive analytics are uniquely suited to understand the nonlinear nature of virtualised datacenter environments.
These advanced analytic approaches enable enterprises to combine IT and non-IT metrics in such a powerful way that the data generated by the networked computing stack can become the basis for automated and embedded business intelligence. In the most sophisticated scenarios, analytics and machine algorithms can be applied in such a way that the datacentre learns from itself and generates insight and models for decision-making approaching the level of artificial intelligence.