| This week's sponsor is PGi. |  | Webinar: IT and Marketing: Extreme Collaboration
Tuesday, August 26th, 2pm ET / 11am PT | New Editorial Event!
Media outlets love to focus on the tension between IT and marketing. But if it's a war, both sides lose. Instead, CIOs have to partner with CMOs to help deliver on aggressive business goals in an ever-changing landscape. Register Today! |
Editor's Corner: A death of tact Also Noted: Spotlight On... Why make Microsoft pick whether to violate EU or US law?
Australia may push out its cloud; SUSE pushes an IaaS cloud; and much more...  Follow FierceEnterpriseCommunications on Twitter!
News From the Fierce Network:
1. Leading your IT team through a slump or setback
2. Scientists lured by freedom to experiment beyond the hallowed venues
3. 4Q predictions by co-author of Big Data for Dummies
| 
A death of tact
 "No words."
This was the entire message posted to Twitter by both Billy Crystal and Whoopi Goldberg in response to their having learned of the suicide death of their friend and frequent colleague, Robin Williams. As genuine human beings, they both realize when silence is the most appropriate reaction. On Wednesday, The Washington Post reported that Zelda Williams, Robin's 25-year-old daughter and herself an actress, declared a moratorium from Twitter and later Instagram after having received disparaging, dismaying and bullying comments from some individuals. Some text was said to have accused her of complicity in, or responsibility for, her father's death; and the description of the pictures she saw does not deserve reinterpretation from me. This publication is about communications. As one that focuses in larger measure on enterprise communications, it must also be about tact: the way in which we present ourselves to the world over a communications medium. So be advised that I am not using my column this week to take a break from this broader topic and discuss the disease of ill behavior that afflicts our society. Indeed, I don't believe in such a thing. Quantitatively speaking, the evidence is not there. As Thursday's Post follow-up makes clear, a great deal of the abusive Twitter comments came from just two people. This column, like all my others, is about technology and the way it is utilized by more than two people. Amid the larger unfair disservice to which Zelda was a victim this week, she has had to come face-to-face with what technology has empowered certain small groups of small people to do. As an equalizing force, technology has elevated the power of the small to a level where they can do greater damage, which is dangerous when damage is all they are capable of doing. As I said online in 1995, the week that Timothy McVeigh bombed a federal building in my old hometown, technology has exponentialized the power of the small. In my intense sorrow, I said it makes the wrong people equal to us. And I caught flak for it, being told it was not the appropriate response. Amid our genuine desire to provide the appropriate response for people who could truly use a shoulder instead of an aphorism, many of us end up saying very little. For someone who talks so much and writes so long, a colleague told me recently, I don't have much of a Twitter feed. I responded that, for some reason, while a lot of what I say in 1,200 words is said to be "on the mark", someone almost always complains that whatever I say in 140 characters is "too much." Which leads me to Ferguson, Missouri. Someone in a position of public power will inevitably say, probably next week, that Ferguson reminds us we need a public conversation about race in America. It's a message that condenses well. Yet the medium that would support such a conversation is clogged, right this very moment, with individuals responding to exactly that message--"Hey, folks, we need to talk about this"--by saying it's inappropriate to accuse folks, or accuse them, of racism. The echo chamber of a so-called social network is turning out to be a poor medium for the expression of human feelings. There are things one would never say to another human being in close company during a time of grief, even if someplace within one's potentially twisted mind, they're being thought. It isn't for fear of being publicly condemned. Public condemnation constitutes the sum total of some folks' content streams, both back and forth. We restrain ourselves because we can see each other's eyes. Human contact is high-bandwidth communication. In close company, we may permit ourselves the right to express feelings we cannot explain intellectually. In public, "politically correct" discourse may be so devoid of heartfelt personal opinion that it reads like the back of an aspirin bottle. The role of any editor is to find the appropriate place in-between. When I learned Robin's death was a suicide, I was furious. I did not tweet what I felt. Had I been a cable news anchor, my feelings might have inadvertently filled the interminable moments of airtime between actual information. And I may very well have said something I would regret. We censor our feelings for public consumption, as we should, unless we have already elevated the value of every expression we make to that of "live". In the moments before the late Peter Jennings completed an impossibly long on-air shift on September 12, 2001, Ted Koppel asked him how he possibly fulfills the role of interpreting the mad stream of data pouring in as cohesive information. In a pure moment of honest frustration, Jennings responded, in essence, that he did no such thing, that it was impossible. When feelings overwhelm our intellect, or when our intellect is so besieged by the cruelty of facts that it cannot sustain the stress any more, what sometimes gets expressed is pure truth. That pure truth can expose the integrity of human beings, as it did with Jennings, and as it has with the people of Ferguson for whom the last unjustified police killing was the last straw. Pure truth can also expose the cancer in the souls of those who tortured Zelda online this week. It isn't that Twitter isn't pure truth. It isn't refined truth. High-bandwidth communication compels us to be better people. Low-bandwidth communication, by filtering out any source of feedback from the eyes and the souls of the recipients of our vitriol, permits the raw ingredients of our emotions to proceed unfiltered. At the same time, its narrow aperture disallows any truly beneficial expression of our feelings. There is plenty of room for repeated four-letter words in 140 characters, and no room for one sentence of true sympathy. What's left is best described by Billy and Whoopi. I will be grateful when the system of smoke signals that passes for social communication is replaced with full-motion, high-resolution video. I want the granularity to be so clear that we feel the sweat of the reporter who's slammed against a vending machine by armed police for practicing journalism. I want the resolution to be so crisp, we see the reflection of our faces in Zelda Williams' tears. -SF3 Read more about: Twitter
back to top |
|
Today's Top News 1. Brocade's Tom Nadeau: OpenStack deserves a champion
It is not only fair to say that Tom Nadeau wrote the book on software-defined networking, it is entirely accurate. His definitive text on the subject was published by O'Reilly a year ago. At Cisco, Nadeau was the principal architect for most of the management information bases (MIB) for Multiprotocol Label Switching protocol (MPLS), and several of the other principal network routing protocols implemented in all the world's networks. He remains a long-time contributor to the IETF, co-chairing the Network Modeling and Energy Management working groups. And last February, he was hired by Brocade Networks, Cisco's biggest competitor in the network component space. Brocade has the most to gain from SDN's and NFV's success, even if Cisco is successful as well, because it would even the odds. It could even lead to parity. In an interview with FierceEnterpriseCommunications Wednesday, Nadeau shared his expert opinion on why open source development is the proper model for achieving parity in the field he helped launch. In the course of our talk, he shared some compelling viewpoints on the two platforms upon whose success Brocade clearly depends: the OpenStack orchestration layer, and the OpenDaylight SDN/NFV layer. The first danger Nadeau perceives with the deployment of proprietary controllers in the SDN space is the obvious one: VMware's recently acquired Nicira, Juniper Networks' Contrail, and Alcatel-Lucent's Nuage all help guarantee some degree of vendor lock-in, he argued. But the second danger, in his opinion, lies beneath the surface: "They're inherently fragile systems," he tells us, "because they're closed and proprietary code. Because you can't see the code, it's a black box, so you need to try to test every permutation and combination of how that thing will operate, without actually being able to look inside." Up until recently, OpenStack had been presenting itself as the product of a team of development comrades who put their loyalties to their employers aside to produce a reference platform that everyone can use in equal measure. But that was before the arrival of HP into the market. Then suddenly, there was serious talk about whether HP and rivals Red Hat and IBM were openly (to coin a phrase) jockeying for the very leadership role OpenStack claimed not to need. When I put the issue directly before Brocade's Tom Nadeau, he said he felt OpenStack had reached the point in its evolution where a champion of the technology is actually required. "If you look at any successful open source project, you will find a champion behind it--a champion being, not just somebody who sells it commercially," he remarks. "The Linux kernel is a great example: Linux itself would not have become a useful, enterprise-class, robust implementation without significant people typing on keyboards. When IBM decided to change from its proprietary version to go all-in on open source, that was a game changer for the industry. They contributed hundreds of developers to working on it, and that, in part, is why it's successful today." A small project might start with five to eight developers contributing part-time work, Nadeau goes on. "But to really challenge industrial-strength, deployable projects, you really need people working on these things full-time, and often in large numbers, because there's a lot of code to be written. So without a champion--I'm not saying it's impossible, but it's a very steep hill to climb." But there's a difference between a champion and a dominant player. Nadeau goes on to say that, in many organizations that produce open source reference platforms, there's one member that's perceived as the "giant", counter-balanced by an abundance of other, smaller players. That coalition may help serve as an opposition party of sorts. In the case of OpenDaylight, the giant had been Cisco. "Brocade came in, made some significant contributions, totally committed to working on the project," he continues. "And what's happened is, a lot of the smaller guys have joined us--'Hey, you guys have straightened this thing out'--and people are now really invested in working on the code. At the same time, it forces that dominant player to play ball, because they'll look really bad. It's a community effort, and if you're seen as anti-social, that's counter to why you're doing open source. Open source is a community effort, despite what some people might tell you." For more:
- SDN: Software Defined Networks by Tom Nadeau, 2013, on shop.oreilly.com
- MPLS Network Management: MIBs, Tools, and Techniques by Tom Nadeau, 2002, on Elsevier.com
- OpenStack doesn't need a leader, it just needs to evolve [by Mike Wheatley, SiliconAngle, January 28, 2014] Related Articles:
Successful proof-of-concept of migration orchestration across clouds
HP to battle Red Hat, IBM in OpenStack space, bottling up Microsoft Read more about: OpenStack
back to top |
2. Intel acquires network chip maker Axxia from Avago, buys itself some time
Late Wednesday, news came from Intel Corp. that it has entered into an agreement with Singapore-based Avago Technologies to purchase the networking business Avago acquired from its own just-completed acquisition of LSI Corp., in a $650 million cash transaction. LSI is the company once known as "LSI Logic," and in recent years, it referred to itself as "LSI Semiconductor Solutions". To understand just what it is Intel is buying (because the end will come as a shocker), you have to wade back through a very long history of mergers, separations and randomly-generated names. Way, way back in 1961, Hewlett-Packard spun off a division for the production of diodes using interesting substances like germanium and gallium arsenide. HP Associates became one of the nation's great producers of LEDs. But in 1999, in one of the many strategic realignments and contortions HP would have to undergo in order to fix a string of "innovations" for which its former CEOs still, at dinner parties, receive credit, HP spun off this business unit to form Agilent Technologies. Write this down, because a lot of things begin with "A" in this story. Agilent ends up going into business with Philips Lighting to produce LEDs. Its semiconductor properties were sold in 2005 to investment firms KKR and Silver Lake Partners in a $2.66 billion deal. The result was the world's largest privately-held semiconductor firm at the time, Avago. Meanwhile, on the other end of history, you'll recall that the reformation of the original AT&T Corp. into a pure-play communications company, resulted in the spinoff of Lucent Technologies. Bell Labs would be part of Lucent, and would also later become merged into Alcatel. But the part of Bell Labs that had been known as Western Electric, and that used to make telephones, was spun off in 2000 to form Avaya Communications. It's now privately held by TPG Capital and, curiously enough, Silver Lake Partners. And the part of Bell Labs that developed semiconductor communications was spun off to form Agere Systems. (See what I mean?) In 2007, Agere merged with LSI Logic to form LSI Corporation. And just last February, in a deal touted as the convergence of 150 years of innovation in semiconductors, LSI merged with Avago. Not with Avaya, with Avago. Get it straight. So the semiconductor arms of HP and of Bell Labs are, at this point in the story, one unit. For the last few years, LSI has been producing a line of communications processors called Axxia. You can imagine the wealth of design expertise and manufacturing wisdom from a producer with such a storied history, culminating in the production of Axxia Enterprise Communication Processors. And you, like I, will probably imagine wrong. Axxia's top-of-the-line model AXE4500 is based on ARM technology, not anything it gained through mergers or acquisitions. And its midrange AXE3500 and starter AXE3400 are repackaged PowerPC components from IBM. Why would Intel possibly want a company that basically repackages and resells its two chief competitors' parts? One possibility is that it eliminates one of the world's producers of ARM and PowerPC products in networking. Anyone who seriously believes Intel would go on producing a networking component with an IBM part in peace and harmony, need only read yesterday's blog post from Intel's Rose Schooler with a magnifying glass: "To enable this vision, we believe that running the four critical network workloads (application, control, data and signal processing) on a single architecture provides benefits by reducing development complexity and risk," Schooler writes (although I've added boldface). "Intel's goal is to offer a common architecture approach that scales up and down at many different price, power and performance points." There's no way Intel will ever put two platforms in one of its enterprise products. So does all this merger and convergence and synergy merely lead to what a chip designer might call a "terminal node?" Maybe not. One clue as to what Intel really bought--and what it really needs--comes from this LSI white paper (.pdf) produced just prior to the completion of its Avago merger, so not long ago. It's a position paper on SDN, and it notes that the plurality of parallel projects to standardize network virtualization is making standardization difficult. "LSI proposes the use of intelligent silicon for a new scalable alternative to solve emerging complexities with SDN protocols such as OpenFlow and, in general, the dynamic nature of software defined orchestration overlays," the white paper reads. "As SDN is deployed, it is inevitable that datacenters will operate with hybrid networks with some OpenFlow and non-OpenFlow traffic. It is important for the line cards to evolve towards this hybrid state to support OpenFlow traffic gracefully. LSI proposes utilizing flexible Axxia acceleration engines to solve the hybrid network element both for data plane acceleration in addition to performing intelligent control plane acceleration." This is what's worth $650 million, and maybe that's a bargain. If Intel can move Axxia's SDN property onto Intel silicon, then it can produce a complete SDN server platform the same way Centrino was a complete network client platform. For more:
- Intel's announcement of its Axxia purchase by Rose Schooler
- LSI's SDN Strategy and Solution white paper (.pdf) Related Articles:
My take on Intel's near-future support for open standards
Intel: SDN is a 'long-term play', NFV a 'big deal' Read more about: Intel
back to top |
3. IPv4 size barrier for routing tables may be hit this week
One of the long-threatened repercussions of the Internet remaining largely dependent upon IPv4 routing has apparently begun. The number of available Internet routes from IP host to IP host, by many measures, may have exceeded the ability for bits in IPv4 routing tables to represent such a number. And the behavior being exhibited in recent days at some levels indicated that the proverbial cup already overfloweth. "I know this isn't much help, but there are major problems with multiple ISPs since around 4-5 AM EST," reported a Boston-based security engineer Tuesday, to a mailing list devoted to core network outages. "I really don't have much detail, but I have sites that are unreachable from some providers. Looks like Comcast, level3, ATT, cogent, etc." It's one of a growing number of indications that routers throughout the network are incapable of processing IP address prefixes beyond the 19-bit variety, despite workarounds and product fixes recently released by Cisco and others. It's not a security vulnerability yet, and in a blog post Wednesday by Renesys CTO Jim Cowie, he says the Internet as a whole should not be perceived as threatened. But engineers can't escape the annoyance at this point. "The problem is real," writes Cowie, "and we still haven't seen the full effects, because most of the Internet hasn't yet experienced the conditions that could cause problems for under-provisioned equipment... The real test, when large providers commonly believe that the Internet contains 512K routes, and pass that along to all their customers as a consensus representation of Internet structure, will start later this week, and will be felt nearly everywhere by the end of next week." 512K, or 219 (524,288), is the technical upper bounds of IP prefixes held in the routing tables of certain routers, including from Cisco. Specifically, they're held in memory because that's the fastest way to access entries in a database. Routers must contain routing tables in their entirety; they can't divide and conquer or employ hashing, the way CPUs handle memory caches, because the Internet cannot afford the latencies incurred from "misses" the way CPUs can. From a standpoint of difficulty, addressing the problem is a simple enough matter: add a bit to the address field of the routing table, for 220 (1,048,576) addresses. At this point in time, it's the only long-term fix there may be for routers that start dropping huge chunks of IP routes from memory as though they never existed. "No matter who provides your networking equipment, it needs to be able to manage the ongoing growth of the Internet routing table," warned Cisco engineer Omar Santos last May. "We recommend confirming and addressing any possible impacts for all devices in your network, not just those provided by Cisco." Cisco routers, wrote Santos, employ very high-speed memory that it calls Ternary Content Addressable Memory (TCAM). He goes on, "The possibility of TCAM resource exhaustion at 512K routes is a known issue that we all know has been coming for some time. There is no related security vulnerability, and it cannot be easily triggered by a remote, untrusted user." The CIDR Report is one of the Web's earliest and most reliable online tools for measuring the "size of the net;" during the boom years of the '90s, it was used by ISPs trying to boast about their relative size. A Wednesday snapshot by CIDR of the global routing table showed 512,521 active IP prefixes--not quite 512K. Fixing the IPv4 routing problem comes at a cost each and every time. First, there's the obvious performance problem: it takes more memory, and potentially more memory cycles, to fetch an address that's one bit wider. Second, as Renesys' Cowie explains, that memory has to come from somewhere, and typically that somewhere is the unused portion of the IPv6 address table, which is presently underutilized. At the Internet's present rate of growth, we may encounter having to add another bit to the IPv4 table in as soon as five years' time. For more:
- see Renesys blog
- read the Cisco blog Related Articles:
Microsoft: No more US IPv4 addresses left for Azure users
Cisco CEO Chambers: Most corporations will die unless networks can adapt Read more about: Jim Cowie, Cisco
back to top |
4. After history's most public interoperability failure, now US Gov't writes a 'Playbook'
The catastrophe that was the public rollout earlier this year of Healthcare.gov, the website for U.S. citizens' now-mandatory enrollment in public healthcare plans, delighted only the few who thrive on the political theatre of their opponents' visions going up in smoke. For the rest, it was such a testament to government's inability to manage a 21st century project that the GAO's post-mortem report diagnosing the cause of its failure shocked no one at all and thus got little attention through Google News. In the wake of this failure, the White House Monday announced it had created yet another government agency, the U.S. Digital Service. Its purpose, the White House says, will be to apply a new set of best practices and guidelines for the online delivery of government services, across all the other agencies. Mikey Dickerson, the former Google reliability engineer brought in as cleanup czar for the failed first iteration of Healthcare.gov, was named administrator of the new service. What the announcement did not yet explain was the capacity in which Dickerson's observations will be applied by the other agencies--in particular, by intelligence services, the IRS and the Dept. of Veterans Affairs. But a scan of the prototype of the initial best practices list also released this week, called the U.S. Digital Services Playbook, calls to mind another inspired document from an earlier era in our history in how it identifies the responsible party for future government services--a document whose first word is also "We". "We must begin digital projects by exploring and pinpointing the needs of the people who will use the service, and the ways in which the service will fit into their lives," the Playbook begins. "Whether the users are members of the public or government employees, policy makers must include real people in their design process from the very beginning. The needs of people--not constraints of government structures or silos--should drive technical and design decisions. We need to continually test the products we build with real people to keep us honest about what is important." Had the Playbook ended there, half or more of its mission would already have been complete. A best practices guide to the renovation of Congress would not read much differently. The Playbook goes go on to suggest that the entire user experience of government services, however they're delivered--including on the phone and in person--should be modeled and prototyped in their entirety first. A design guide should follow from this model, identifying the potential pain points that users will have, since the necessity of certain government services derives from pain. Then the product should be developed iteratively, using "agile methodologies" (small "a," though veteran software developers know what this really means), with an initial "minimum viable product" (MVP) delivered "as soon as possible, and not longer than three months from the beginning" of the project. The word "beta" is finally introduced into the government's lexicon. Finally, government services should be openly designed, the document says, to be reusable not only by other services but by the public at large. It's difficult to imagine the first Healthcare.gov serving as a model for anything beyond a negative example. But if anything great does come of that example, it may yet be a re-steering of the ship of state toward a positive course. For more:
- see the U.S. Digital Services Playbook
- see the GAO report Ineffective Planning and Oversight Practices Underscore the Need for Improved Contract Management [GAO report, July 30, 2014] Related Articles:
Healthcare.gov security problems not confined to public sector [FierceITSecurity]
White House looking to create new IT A-Team [FierceCIO] Read more about: healthcare.gov
back to top |
Also Noted SPOTLIGHT ON... Why make Microsoft pick whether to violate EU or US law? It should never come as a surprise that genuine journalism appears under the byline of Wayne Rash, the truest of technology journalists. For eWeek over the weekend, Wayne presents exclusive news that the European Commission is considering approaching the U.S. Justice Dept. about its concerns over a U.S. court ruling forcing Microsoft to turn over data on one of its customers--data that had been stored on one of its Ireland-based servers. Doing that would expressly violate E.U. law, and could get Microsoft's service shut down there. Supposedly, treaties between the two government bodies enable the Justice Dept. to seek at least the permission for what it's looking for, from the E.C. directly. But as Wayne's reporting points out, DOJ doesn't appear to have even tried. Read more: European Union Expresses Concern About Microsoft E-mail Privacy Case [by Wayne Rash, eWeek] > Consumerization and the CIO - Now Available On-Demand From devices to services to apps, end users have a lot of choices - and those choices are bleeding into enterprise IT faster than ever. How do these changes affect IT strategy, budget and infrastructure? Register to watch now! > IT and Marketing: Extreme Collaboration - Tuesday, August 26th / 2pm ET / 11am PT Media outlets love to focus on the tension between IT and marketing. But if it's a war, both sides lose. Instead, CIOs have to partner with CMOs to help deliver on aggressive business goals in an ever-changing landscape. Register Today! > Advancing the federal cybersecurity workforce - Wednesday, September 10th | 2pmET/11amPT Join NIST and NICE leaders as they explore The National Cybersecurity Workforce Framework, innovative spproaches to cybersecurity training and workforce improvement, the broader focus of NICE in advancing cybersecurity awareness nationawide, and more! Register Today!
| > eBook: 5 Key Strategies for Successful Mobile Engagement Read this eBook to discover how you can deliver highly targeted, personalized content and services to your customers across all mobile channels – and the key strategies that are critical to a successful mobile approach. Download today! > Whitepaper: Supporting VDIs and Thin Clients Companies have already begun deploying VDIs and thin clients (like Google's Chromebook) on a massive scale. The low-cost, easily deployed workstations present a significant cost savings for companies, but require unique tools to support them. This whitepaper, written by Proxy Networks, outlines the best way to do that. Download now. > eBook: eBrief | Making BYOD Work: 4 Critical Strategies for Midmarket and SMB Companies Bring-your-own-device (BYOD) can be a blessing for mid-size and small businesses. But getting the real payoff requires some attention to details that may differ from those at large enterprises. Download this eBrief to get more practical advice for making BYOD work. | |
Editor: 
No comments:
Post a Comment
Keep a civil tongue.