Too Big to Succeed

Cyprus is in the news as the world wrestles with insolvent banks in that country. The challenge is that a failure of the Cypriot banks would lead to wider failures across the world. The underlying idea is that banks can be too big to fail. The failure of one large bank takes all its investors, bond holders, and depositors down the drain with it, and if this group is large enough it could collapse entire economies. As governments fall with economies, governments declare banks “too big to fail” and find a way to bail out these banks.

So far no government has declared a technology company “too big to fail”, and these tech giants do fail as innovation and competition change the technology landscape. Many tech giants are in the news as they face the latest shift to touch screens, cloud, and big data. Should these firms be declared “too big to fail”, or are they simply too big to succeed?

A core competence is the result of a specific set of skills or production techniques that deliver value to the customer. Such competences enable an organization to access a wide variety of markets. Executives should estimate the future challenges and opportunities of the business in order to stay on top of the game in varying situations. -wiki

The key question: when is giant size helpful in technology? A standard way of answering this question is to identify the core competencies of a firm, and testing whether the firm’s activities focus on the competencies. The standard alternatives to focus are vertical integration or a conglomerate of unrelated parts. A firm which focuses on core competencies will succeed, others will not.

Apple has always been unique among technology firms, choosing to build both hardware and software of its own while “Wintel” took a commanding market share. For much of its existence Apple fit the idea that a preference for vertical integration rather than focus on core competencies leads to weak performance. The stunning move into portable music players – the iPods — certainly changed results for Apple. Vertical integration was retained with this new product, Apple still controls both hardware and software. How did Apple succeed with such a new and unrelated focus on music? Perhaps because for Apple a focus on music was not unrelated to its core competency . Apple long held market share for students and artistic computer users, its college student customers wanted a better way to collect, share, and bring music with them (Napster anyone?). Apple used its competency with these consumers to change the user experience and redefine technology at the point of use. Expansion of the music player to include games and other Apps, touch screens, cameras, phones followed by increased device sizes of tablet computers – iPads – follow a focused core competency in consumer computing.

HP provides a counterexample. Known for printers and servers (including consolidation of DEC and Compaq), HP decided to vertically integrate to capture a bigger share of corporate IT spending. Moving first to consulting (EDS acquired) then to networks (3Com), tablets (Palm) and lately to software (Autonomy) HP has yet to find a way to succeed as a whole. HP is a giant, and it wants to get bigger like IBM, but without identifying core competencies it will remain a conglomerate that is too big to succeed.

Amazon provides an example of growth through focus. Started as an electronic bookseller it has leveraged its core competencies in electronic retailing and order fulfillment to expand to all retail products. Amazon goes further by hosting operations and electronic stores for other retailers. Even its foray into tablet hardware – Kindle — and e-books center on the core competency.

Oracle dominated in databases then vertically integrated into applications. Acquisitions drive growth and substitute for innovation culminating in Oracle’s entry into hardware (Sun acquisition). Meanwhile customers move on with Hadoop and NoSQL instead of Oracle’s core database products. The mighty Oracle sales engine is faltering as distractions replace the “value to the customer” component of core competencies. Can Ellison regain focus or will Oracle acquire itself into too big to succeed?

In a world where the giants lose their way, what is a smaller tech firm to do? Get very familiar with your core competencies and focus your strategic moves in those areas. Your customers define the value of your competencies. If you are vigilant your customers may lead you to unexpected value as Apple moving into music. They will certainly lead you to consistent value as at Amazon in all things retail. Chasing size for its own sake may separate you from your customers as at HP and Oracle – then you risk becoming too big to succeed.

Posted in Uncategorized | Comments Off on Too Big to Succeed

Implementation Risk

Another day, another ERP vendor sued over an implementation failure. The customer did not achieve success implementing core ERP software and naturally blamed the vendor for inadequate support. You would think that by now someone somewhere would understand what it takes to successfully implement ERP software – and you would be correct. Success depends upon the executive management of the organization implementing an ERP system for its own use. Only the executives can decide which features are priority number one for an implementation. Only the executives can say no to users who wish to alter commercial ERP systems rather than alter their personal workflows to take advantage of purchased software. Only executives can monitor the implementation process to see that it is on time and on budget. (The implementation process and executive responsibilities are detailed in HarrisData’s Quick Start Implementation Guide). Unfortunately, some executives would rather shift blame than do their jobs.

That said, no vendor should hide behind this executive responsibility instead of doing their part in assisting a customer organization’s implementation process succeed. In this instance, the vendor faced several self-imposed challenges which may have inhibited their ability to assist the customer. The ERP software was sold through a distributor, putting an intermediary between the vendor and direct communication with the customer’s executives. The ERP vendor was acquired by a larger consolidator, requiring new relationships between the vendor, the distributor, and the customer. The vendor’s project team assigned to the customer implementation left the vendor (this is attributed to the acquisition, but occasionally happens anyway). It is the vendor’s responsibility to support the customer through these day to day challenges (although the vendor in question may focus acquisitions more on big footprint / cost reduction than improved support / organic growth of acquired products). The vendor did not do so to the customer’s satisfaction.

Frank Scavo focuses on concrete things a customer organization can do to minimize these type of implementation risk. They are an excellent set of recommendations. HarrisData makes it easy for our customers to take the recommended steps by implementing our Software Customer Bill of Rights in our OmniLicense for ERP software. We also try to maintain contact with customer executives throughout the implementation to remind them of their responsibilities and options, and to make sure HarrisData lives up to our commitments. However , success depends upon the customer executives and is their accomplishment when achieved.

Posted in Software Support | Tagged | Comments Off on Implementation Risk

Surviving the Drop in Business Investment

Truly amazing news as the UK’s National Museum of Computing keeps the WITCH online — mechanical relays, paper tape and all!

The computer’s “flashing lights and clattering printers and readers provides an awe-inspiring display for visiting school groups and the general public keen to learn about our rich computer heritage,” the museum says.

After 61 years, it can still store 40 8 digit numbers and multiply two of them in a blazing 5 or 10 seconds.

Unfortunately, in the uncertain economy we face today business executives must consider keeping older (not older than WITCH, just old enough to be behind the technology innovation curve, ed.) systems running their IT processes for the next year or so as capital investment budgets shrink. This fact will challenge technology firms as new sales fall off, moreso should executives elect to forego software maintenance renewals. What should technology executives do to meet this challenge?

When static business investment is the optimistic forecast, so is static revenue. Stasis is not good for technology revenues and lower revenues can lead to reduced innovation budgets and more stasis. Skip over the traditional “rightsizing” and “acquire bigger footprint” financial steps all firms will use, and consider instead what positive steps may be taken to assist the business executives in gaining increased value from older systems (thus strengthening maintenance renewals and possibly leading to new licenses ). The most positive steps technology firms can take focus on maintaining or increasing innovation … then delivering the innovation to customers without excessive costs.

The greatest opportunity comes from integrating communications about customer issues, while automating workflows to turn customer issues into product innovations. The idea is simple – too many technology firms have islands of information related to customer problems. From implementation team notes, to help desk records, to product quality tests, to requirements and design documents in R&D the core data that should be driving innovation is scattered, missing where needed, and unused. Technology industry executives know how to recognize this problem, we solve it for our customers every day. The answer lies in unifying innovation related work flows through a single shared database of customer issues and resolutions.

Once the innovation data is unified, it is time to analyze it. Complex analysis is another capability technology firms provide their customers that should be turned inward today. Identify issues experienced by many customers. Rank them by cost / opportunity to each customer. Rank them by total cost / opportunity to all customers. Compare the ranks to the cost and time required to deliver an innovation targeted at the issue. Deliver the low hanging fruit (low cost to you, high value to your customers) quickly. Identify customers who might be experiencing the issues your new innovation addresses – use outreach to help them benefit from your work as well. Lather, rinse, repeat. Customers who receive a continuing flow of business benefits are more likely to renew software maintenance – both the technology vendor and the customer can point to concrete examples of the value underlying the expense.

Technology executives need to recognize the budget constraints facing our customer executives. Next we need to apply our knowledge and skills to our internal innovation engines and speed innovation to our customers. Those who can deliver innovation to customers within the framework of the software maintenance agreement (i.e. at minimal to no extra charge) will experience stronger maintenance renewal rates. Strong maintenance renewal rates allow technology firms to sustain innovation budgets and are the key to surviving in this economy.

Posted in Software Support | Tagged , , | Comments Off on Surviving the Drop in Business Investment

Tablets Mean Changes for the Enterprise

Plenty of hype has been offered as tablet computers (first the iPhone, then the iPad) arrived, sold in enormous numbers, and slowed PC sales. Software vendors were quick to deliver Apps – partly to ride the new wave and partly to provide true ease of use and mobility for selected bits of their enterprise applications. It is finally dawning on some that tablets mean something different than what has been hyped. At HarrisData we believe tablets usher in a new era of consumerization of enterprise computing where the users act as consumers in selecting devices, and expect corporate applications to match the ease of use found in consumer Apps. Joshua Greenbaum notes Microsoft had a similar realization and provided its first answer in Windows 8.

The real tablet revolution was about multi-touch, not mobility, and with that tactile experience we’ve discovered a new way to improve the user experience that, while it found its first mainstream expression on the tablet and smart phone, need not be limited to those devices.

The enterprise focus must change to incorporate the user experience in the same way that consumer products focus on buyer experience. The big question is how does the vendor accomplish this focus? Creating a constellation of Apps around the corporate system is the most common path taken. However this creates a bifurcated user experience where multiple devices are required to accomplish normal workloads – an App on a smart phone or tablet for trivial matters and a PC for heavy lifting. HarrisData chose a different path, eschewing Apps for a touch screen enhanced browser interface that works both on a tablet or the PC with a single App-like user experience on all devices. Microsoft chose the HarrisData path.

I can access virtually all the functionality I need from a web page, instead of running a specialized app to perform the task. No need for an app to check public transit, the weather, airplane schedules, my Amazon account, listen to my local PBS station, check out Twitter, Box, or any other service I have a subscription to. I just point a browser – or better yet a bookmark – and away I go.

Corporate applications built on this framework deliver the promise of the tablet without sacrificing the power of current PC functions like word processing or spreadsheets. Creating spot Apps for popular functions fails to deliver.

It’s not enough to have your vendor promise a killer iPad app to satisfy your mobile user fan base. You should be demanding, and the vendors should be providing, multi-touch-based enterprise apps that can live in the mobile and the desktop world interchangeably.

Of course, once the interface is delivered there is much more work to be done. Underlying corporate applications must be restructured to capture and deliver the new work flows enabled by typical tablet Apps. HarrisData is currently restructuring our first applications with resulting products to be delivered soon.

Posted in Cloud computing | Tagged , , | Comments Off on Tablets Mean Changes for the Enterprise

The Pendulum Swings

Anecdotal evidence of a US manufacturing rebound has been building for the past two years. Here and there manufacturers “reshored” production to address problems encountered in the rush to low labor costs in China and elsewhere. New statistical evidence supports the anecdotes with strong evidence of a trend towards a greater production share for the US.

Manufacturing’s share of GDP rose in 2011 to 12.2 percent from 11 percent in 2009, the best streak since the ’50s. During the first half of 2012 the U.S. factory sector grew by another 3.1 percent versus a 4.8-percent global decline.

This year manufacturing grew in the US compared to a decline globally. The reasons have to do with US strengths as much as off shore challenges.

The Chief Executive article highlights several strengths:

    American plants have become vastly more productive.

    American advantages in technological innovation are becoming more important as manufacturing becomes more complex. Smart automation, lightweight materials, nanomanufacturing and rapid prototyping—all are areas where U.S. companies excel.

    As American consumers have become pickier, demanding more options in the goods they buy, it’s easier to satisfy them by building close to them.

Each of these strengths is supported by improvements in manufacturing ERP systems. Richer functionality combines with deeper integration to eliminate administrative delays in processing configured orders to satisfy customer demanded options – it is now possible to release orders for customer selected options directly to the factory floor. In this order management environment combined with first time quality improvements the automation reduces delays to physical production and transport times.

Off shore supply chains take too much time for transport, and have provided iffy quality.

Darlene Miller, owner of Permac Industries, a high-precision machine shop, is one. In 2006, she hired a Chinese contractor to make machine blades to exacting specs. But each shipment arrived flawed, with some blades not uniform, made of the wrong material or falsely labeled “certified.” Complicating matters, she had to buy a year’s worth of blades at a time.

For Dan Shimek, owner of the Outdoor GreatRoom Co., offshoring became too inefficient. Shimek used to buy all his wooden pergolas from China and fiberglass fire pits and pergolas from India, but minor problems on hardware orders created major issues because reshipping took six weeks. And sometimes reorders on a hot product arrived too late, costing him a season’s worth of sales.

The lower price of off shore production cannot compete with the business speed, flexibility and reduced overhead of production in the US. The reshoring trend is based on sound fundamentals and will continue in the face of economic headwinds.

Posted in Uncategorized | Tagged , | Comments Off on The Pendulum Swings

Storm vs Cloud

Once again, Mother Nature dashes the best laid plans of men. The Friday storm took down Amazon’s cloud data center in Virginia (once again). The storm disrupted power supplies for much of the greater DC area, including Amazon’s US-East-1 region. Apparently the backup power came online too late, and a data center’s worth of virtual servers needed to be restarted. It took Amazon about 24 hours to fully restore cloud services.

This time the cloud crash took down some big names including Netflix, Instagram, and Pinterest. I assume these tech leaders did not choose Amazon’s remote failover capabilities when they set up their cloud based businesses. Simply using the cloud does not insulate anyone from operational problems at data centers. The cloud can be effective provided the service providers and their customers apply normal risk analyses in setting up cloud implementations – remote failover is one response to weather related risks that needs to be considered.

In fairness, it is likely that many on premises data centers were affected by the power outage. These needed to configure their applications for likely risks as well. It’s just that with all the cloud hype, the failures on premises are ignored as are the successes.

Posted in Cloud computing | Tagged | Comments Off on Storm vs Cloud

Hype, Mixed Messages, Snark, and Cover-up

The convergence of cloud, mobility, and consumerization of business technology make a competitive environment for ERP applications even tougher. No vendor can deploy all the answers today. This leads to exaggerated claims and snarky counter claims by the big ERP vendors. Larry Ellison shows why he is the king of controversy with his recent claim of 100 cloud apps.

Larry Ellison claimed in his Oracle Cloud presentation last week. I’ve just checked, and his exact words were “Over 100 enterprise-grade applications running in the cloud.”

Analysts wonder just what Ellison is talking about.

Indeed, Ellison at one point boasted that Oracle was eating its own dog food by running four of its own cloud apps, and a familiar list appeared on the screen: Fusion Sales and Marketing, Fusion Financials, Fusion Talent Management, and Fusion CRM. (When these four apps appeared on the screen, I couldn’t help asking in a Tweet where the other 96 apps were.)

The analyst consensus is that Oracle is creating hype through an ever changing definition of what is and isn’t an app. This hype is kicked into high gear when Ellison claims SAP doesn’t get cloud.

For example, Ellison claims that SAP has done nothing in the cloud except for its acquisition of SuccessFactors, and that it will have nothing otherwise in the cloud until 2020. He conveniently overlooks SAP’s five or seven year effort to develop Business ByDesign, a full-suite multi-tenant cloud ERP system, which SAP has has sold to over 1,000 customers.

Yeah, Ellison is doing his normal thing again. The real question is what value is there for Oracle customers in what has been announced. Frank Scavo finds the cover-up that Ellison’s hyperbole is designed to push behind the curtain.

Back-channel discussions indicate that nearly all Oracle Fusion application sales are for cloud deployment, not on-premises. It appears that this is the case not because Fusion can only run in the cloud (like or Workday) but because Fusion technical requirements are so complex that virtually no organization wants to deploy Fusion Apps on-premises. It is easier to simply turn over the infrastructure and application management activities to Oracle.

The Oracle Fusion Apps are too complicated to run on-premises? This is a major technology fail event hidden behind bluster and snark. The whole argument behind the cloud is that cost are controlled by having the vendor deal with the application. However if it is too complex to run on premises, the vendor may be able to manage the apps from behind the curtain in the cloud – but not at a competitive cost.

HarrisData revealed our approach to consumerized technology at our recent user conference. Yes all the typical checkmarks are included (cloud, mobility, analytics) . However the metric underlying our approach is whether the new consumerized technology is easier to deploy than our prior releases. This ease of deployment is a key software engineering goal of the project. The economic justification for time invested in ease of deployment: lower cost deployment for our customers on premises equals higher ROI for customer projects with an added benefit of lower cost deployment and higher ROI for us behind the curtain in the cloud. Better outcomes for customers always create better outcomes for vendors, it is a virtuous circle. It is simply stunning that other vendors miss this key principle of application development.

Posted in Cloud computing | Tagged , , | Comments Off on Hype, Mixed Messages, Snark, and Cover-up

To Boldly Go Where No One Has Gone Before

[CEO presentation at this year’s user conference, ed.]

Charting a new course
Where to go?
How to get there?

As CEO, my focus is on leading HarrisData and our customers to a successful future. That requires answering a couple basic questions.

Where to go?
Key trends pointing the way:
Smart phones and tablets are changing the way people interact with computers

    Mobility and multiple devices
    “Touch and speak” interfaces
    More iPads shipped than PCs by any single manufacturer
    Fastest adopted technology in human history

“Consumerization” of technology

    There’s no instruction manual for the internet
    There’s no menu for the internet

The most important change facing business application vendors today is the Consumerization of IT. After years of hype about Cloud / Mobile / Apps, the extraordinary adoption of tablet computers and smart phones represents a major change in the way people interact with software. In the past we could count on new hires being familiar with Excel and Word style point and click interfaces and menus. The new hires today know Google, Facebook, and Twitter – applications that do not have menus, that are optimized for touch screen and voice interfaces, and are so intuitive that little training is required to operate them. Further these new hires often operate more than one device at the same time, Twittering on the smart phone, searching on the tablet, and checking Facebook status on their laptop. The actual devices are not selected by corporate IT, but by the user themselves who bring their own device(s) and expect to be able to use them at work. Our challenge is to let our employees be consumers of devices while handling the back end applications and security necessary to a business environment.

How will application users react?
Study enterprise application users to find out how they respond

    Engaged leading user interface design firm projekt202 to help find out
    Watched real people do real tasks using new devices and interfaces

“Surprised and delighted”

With such a fundamental change in the way enterprise applications are organized in front of us, HarrisData engaged the experts at projekt202 to learn what users wanted, how to organize work flows, and how users reacted to the new approach. Several normal people (users of HRIS applications) were asked to participate in the study. We watched from afar (real-time over the internet) as these people attempted to perform normal tasks using the new approach. One by one they were presented a tablet featuring the new software and let at it without any training or explanation. I particularly enjoyed watching one skeptical person rant about how strange and inappropriate the application was, then plunge in, and about 5 minutes later start singing its praises! The results from this study: users were “surprised and delighted” with the new format.

User-centric interface
Most enterprise applications can be improved dramatically using new techniques


New HarrisData Workspace

What is this new format? It is a way of presenting an enterprise application which eschews menus and allows users to organize their work, collaborate with each other, and adapt the workspace to their own needs. Organize, collaborate, and adapt are the three core capabilities that consumer technology does well. Adding these techniques to enterprise software is what Consumerization of IT is all about. The new HarrisData workspace achieves the consumerization of HarrisData’s software.

The RTI Software Division has a workspace for CRM that got us thinking quite a while ago. Note that while it adheres to conventional “Windows” standards, it does provide user centric features. Tabs for task lists, collaborative work flows, and adaptability (via the IT department) are all present. It is a very effective way to manage a variety of tasks that come about when helping people through a support issue or sales cycle. This is a good starting point, but more industrial than consumer oriented.

Applying what we learned the new HarrisData workspace adds capability while achieving consumer orientation. Search, Favorites, History, routine tasks (what you want to do), assigned tasks (what your boss wants you to do), and alerts allow the user to have everything at their fingertips and organize it all as they see fit. Note, not a menu in sight!

How do we get there?
Server-centric architecture

    access from any device (mobile or desktop)
    Better security
    Integration via web services
    Cloud deployment options

Device-independent interface

    Windows, iOS, Android, what’s next
    Rich “touch and speak” clients
    Integration with common client features

How do we get the new HarrisData workspace to you? First we leverage a strong server-centric architecture for enterprise applications. The core processing, security, and data are kept on the server where it is easier to deploy, manage, and control. The server can be on premises or in the cloud. Web services allows extension of enterprise functionality and integration to peripheral systems (such as your e-Commerce site). The client is device independent, it runs in a browser and works with any device you user / consumers choose. At the same time the workspace can access common client device features (e-mail, contact lists, calendars, cameras) as needed. Since consumer devices like smart phones and tablets are the most stolen / lost / replaced items in our world, it is extremely important to keep the data and security on the server side so nothing is lost / and new device setup is a snap.

How do we get there?
Use Industry Standards

Use proven Frameworks

On a technical level, the key is to avoid Apps. Apps are device specific, must be independently developed and maintained, and require user effort to update on the device. Remember, users will bring their own devices anyway, it is our task to see that they are productively utilized. In response, the industry has developed core standards for consumerization which accomplish all that Apps do, but without the hassle. Use of these standards ensures operability across all devices. Proven frameworks including Sencha Touch and PhoneGap provide consistent feel and access to device capabilities. IBM’s Cognos provides delivery of BI and Analytics via consumer devices (it still requires report creation the old fashioned way).

How do we get there?
Build a common interface
Add web services layer to current applications

    Re-architect server processes as needed

Connect interface to current applications

    First project: HRIS
    Next project: CRM
    More to follow

To summarize, HarrisData is developing a consumerized, browser based workspace interface for all our applications. We are adding web services to all our applications, redefining processes where needed. Today we are introducing our new HRIS – a complete payroll and human resources application utilizing everything we have learned and featuring the new HarrisData workspace. Our next project will be the RTI Division’s CRM applications CustomerFirst , SalesFirst, and WebFirst. The rest of our applications will follow.

HarrisData is boldly taking all our applications to the new consumerized IT environment. No other enterprise vendor has done this, although several have dabbled with point apps alongside their applications. Expect “surprised and delighted” users / consumers at your organization in the near future.

Posted in Cloud computing, PHP | Tagged , , | Comments Off on To Boldly Go Where No One Has Gone Before

Cloud, Mobility, and Big Partners

IBM recently announced new PureSystems and PureFlex hardware for inhouse cloud deployment. The good news is, as Frank Scavo notes

For some reason, Oracle, Microsoft, and IBM are able to make life easier for small developers.

Thus HarrisData achieved certification on the PureFlex platform for its entire suite of products with little fuss.

For a comparison, Scavo looks at SAP as it rolls out Mobility technology and encourages its partners to develop apps for customers.

But for a hint of what it’s like for a small developer to work with SAP, take a look at this blog post (on SAP’s own SCN community site!): Why Does SAP Make This So Complicated?

I am pleased HarrisData competes with SAP and partners with IBM.

The other interesting commonality between the IBM and SAP announcements is that Cloud and Mobility are now re-energizing hype by merging into one giant hype machine. Big partners provide lots of development tools and the hype, while small developers provide the apps that actually do anything related to the cloud or mobility. While mobility makes sense in certain applications (mobile payroll or accounting seem odd), and cloud makes sense for some companies and not others (buy vs rent), the combination is presented as more global than either alone. This misses a key point: mobile (now in the form of tablets rather than smart phones) is preferred because the touch based user interface is easier to learn and navigate than mouse based point and click interfaces. Very few modern ERP functions involve lots of keying (thank the growth of e-commerce like EDI, web based sales, and ACH plus scanned employee time and inventory transactions). Many of these functions are a natural fit for the tablet. This is especially true if the mobility is within the walls of the company office, and much expensive PC infrastructure is no longer needed. That the tablet still works outside the walls is a bonus. HarrisData is developing our software to take advantage of this increased ease of use.

Posted in Cloud computing | Tagged , | Comments Off on Cloud, Mobility, and Big Partners

Cloud Security, Full of Holes and Bugs?

Maybe not full of holes, but certainly the Cloud requires us to think carefully about what we are doing. Mel Beckman addresses the top 5 vulnerabilities of the cloud:

Each of the three cloud service models has at minimum one unique security vulnerability, and cloud services in general share a couple of serious security risks. I’ve found five vulnerabilities that are significant because they’re so often overlooked, even when they’re easily addressed.

Mel cites inadequate passwords and the need for multifactor authentication, a problem which hackers are all too ready to exploit. Add the need for encryption of the cloud conversation (including authentication) using something like SSL 3.0. Then there are factors only your cloud vendor can control, but you need to confirm: the virtual server “snapshot problem” where others have access to virtual memory pages including all encryption keys; backup redundancy (does your backup provider back up its own datacenters and how well); and treatment of API keys as data rather than keys to be secured. For the full explanation of each vulnerability and how to address it, please read the full article.

Charles Babcock finds a security bug caused a major cloud service outage at Microsoft.

A process meant to detect failed hardware in Microsoft’s Azure cloud was inadvertently triggered by a Leap Day software bug that set invalid expiration dates for security certificates. The bad certificates caused virtual machine startups to stall, which in turn generated more and more readings of hardware failure until Microsoft had a full-blown service outage on its hands.

The bug was part of a customer workload, reminding us that in the cloud your neighbors’ bad habits affect you. Given the nature of the bug, any regression test from the Y2K days should have captured it. One customer neglected this old standby quality requirement impacting Microsoft and many of its customers.

Art Wittmann puts it all into perspective regarding what to consider before using the cloud, before wrapping your head around the above risks of using the cloud properly and safely.

If you’re a weekend warrior and you need a tile saw, jack hammer, or concrete chainsaw for a small project, you go rent one, even though you know that a single weekend’s rental price is probably a quarter of the price to buy the tool. But if you lay tile for a living, you buy your own tile saw. The point is that for very infrequent use, the price of the rented tool almost doesn’t matter. You need it when you need it, and you don’t want to own it under almost any circumstance. If you have computing needs like that, the cloud is for you.

He identifies very creative use of the cloud for development and proof of concept activities prior to moving production in house, as well as using the cloud for “burst mode” capacity needs for a limited timeframe. The idea is to be flexible and creative about using the cloud, but do not forget about the advantages of on premises for your day to day workloads.

Posted in Cloud computing | Comments Off on Cloud Security, Full of Holes and Bugs?