Monday, December 31, 2007

Innovative Minds Do Not Think Alike

The more knowledge you possess the less you think outside the box. Experts in a field can benefit from an outsider's perspective. This is particularly relevant to the gadgets we make and the software we write. A telling sentence from the article: "It’s why engineers design products ultimately useful only to other engineers."

read more | digg story

NetDirector-enabled Scripts Requirements Document

We've been working with some ND users and prosepcts to refine the requirements for running users' own scripts through NetDirector. We'd love to know what you think!

Objective

Enable NetDirector users to upload, view, edit and run scripts through NetDirector. This provides the ability to schedule script actions, audit actions, and rollback actions.

Features
  • Create New Script:
    • Text editor in ND to create scripts
    • or Pull scripts from another server
    • or browse local disk
    • Specify script name - this will show icon and script name
  • Assign scripts to servers using the standard ND server tree
  • Allow the ability to schedule the script to run using ND's calendar
  • Script action rollback acheived either through:
    • specifying files to backup before running script and restored through backup
    • and/or add a rollback script to undo the action - can write the rollback script in the editor, pull from another server or browse local disk
  • Need an icon that wil appear in the services pane
  • Future feature
    • workflow - specify multiple scripts in sequence

Sunday, December 23, 2007

Centennial Networking Lab

Recently I've been spending some time over at NC State University's Centennial Campus. Designed to be a cooperative University/Industry park, the campus mixes corporate offices (Red Hat's HQ, for example, and ABB's Raleigh office) with University departments and facilities. There's even a small residential area with some really nice (and expensive) all-brick townhomes.

The reason I've been over there lately is because John Bass, who runs both Centennial Networking Lab and the related Centaur Lab (one facility, two missions), has been gracious enough to let a few of our servers call it home.

Upon entering the lab, you see many things you'd expect to find in such a facility - racks of servers and networking gear humming along, air conditioners, raised floors, neatly threaded CAT5 cables, a roomba. Look closer at the racks, though, and it becomes clear that this is no ordinary data center.

Blue-faced Juniper M20s sit along side Cisco GSRs and Catalyst 6500s, all connected to the rest of the world via deceivingly skinny OC-48 and 192 fiber connections. Want to simulate thousands of Web users accessing your web site from their browsers to see what your limit is? Spirent's gear is up to the task. One and a half racks of Dell servers crunch numbers for the NCSU Physics department, modelling what happens right after a star dies.

If one of the lab's tenants needs to test QoS across MPLS, "We can throw up a quick LSP mesh," John says in a decidedly nonchalant, but also completely modest, way that comes from having done it umpteen times.

The lab serves as the Raleigh, NC site for Internet2, which helps explain the tremendous routing capability and fat, redundant OC-48/192 connections.

You may know John's name from his writings in NetworkWorld Magazine, where he frequently publishes the results of the product bake-offs he performs at CNL. I first got to know John back in 2005 when we wanted to test the scalability of NetDirector. John and his team designed the test, implemented it and, in the process, uncovered several areas for improvement. For my money, John is one of the best data center/networking professionals in the business - he knows his stuff inside and out, honors his commitments, is generous with his knowledge and contacts and is a great guy to boot.

Cheers JB - YTM!!

Friday, December 21, 2007

Whitehurst as New Red Hat CEO - A Great Choice

Today, my version of the lyrics from the Beatles song "A Day in the Life " would go a little like this:

"Woke up, got out of bed, jaw hit the floor." Given that the announcement of Szulik stepping down came as a surprise even to many company execs, I guess there's no reason why I should have seen it coming. But, still, WOW! Total shocker.

Reading the reaction from Red Hat watchers has been fascinating, too. Here's a quick rundown.

  • Phil Manchester of El Reg's developer bureau provides a real nice write-up that puts this move in a broader market context. He notes that Whitehurst's appointment can be seen as a big step for open source towards the mainstream, reflecting the shifting "suits to sandals" ratio in open source. I couldn't agree more.
  • CNET's Matt Asay: Asay's first reaction was to throw a conniption, saying, among other weird things, that he'd lost all faith in Red Hat. Then I guess he poured himself a strong one and, an hour and a half later wrote that "he probably overreacted." And in his latest entry on the subject, Asay admits that, because Open Source is an "operations business" (are there any that aren't??), "perhaps - perhaps" Whitehurst is a good choice.
  • ZDNet's Dan Farber with Larry Dignan and David Berlind: These guys note, quite rightly I'd say, that Lou Gerstner, the guy that turned IBM around, was a complete newbie to technology, having previously run RJR, been an exec at Amex, and before that been a partner at McKinsey. One problem with this comparison, though, is that, by most accounts, including the one provided by Red Hat Investor Relations, Red Hat don't need no turning 'round, thank you very much. But, still, IBM's great success as a services business with Gerstner at the helm proves at least that a computer industry outsider can successfully run a computer company. (Note: It is curious that one of the things Whitehurst is praised for is his big role turning around Delta - maybe some people at Red Hat think they *DO* need a turnaround guy...)
To summarize, El Reg gives us a lot of very useful context, Asay kinda loses it then sort of pulls himself together, and ZDNet tells all us open source fanboys to just simmer down - after all, such a move is not unprecedented.

But the title of this entry includes the words "Great Choice," which goes beyond where ZDnet leaves things. Why? Because I think a former old-guard airline COO who had to watch his lunch be eaten by low cost, low price carriers, has precisely the right background to run the leading open source company. He knows what is going to happen to Oracle, Microsoft, SAP, BMC, BEA, etc., etc., because it is exactly what happened to Delta.

So, if you agree with the above, then it answers why Whitehurst, among available airline industry execs, makes sense. But what is it about the airline industry that makes it a particularly good hunting ground for the next Red Hat CEO? The economics of the airline industry are very similar to software and are especially similar to the SaaS model that is coming to dominate the industry. Both are marked by low and declining unitary costs and by high fixed costs. And network effects are prevalent in both. Lastly, Red Hat's customers - not its community of users, contributors and partners, but its paying customers - buy services, not software. Said differently, Red Hat is in the service business. Guess what the biggest service industry in America is? Hospitality, of which Travel represents $703 Billion.

No, Whitehurst is a very good choice for next Red Hat CEO. Hats off.

Friday, December 14, 2007

Ubuntu's DIY Server Ads - Love 'em or hate 'em?

If you read the same online journals as I do, then you've no doubt seen the new Ubuntu Server ads. I've seen two so far. The one of the guy in the park popping a tent and this one:





























































So, I could be wrong, but I think I saw some of the people in these ads in Boston at the recent Ubuntu Developer Summit. If I am correct that these ads were in fact shot at Ubuntu's offices and that these are Canonical employees, then this would mark a pretty big departure from the traditional, professionally-developed and expensive online ads that usually run on web sites like eWeek. I'm not saying that's a bad thing. In fact, there's something very cool and refreshing about the Canonical ads.

But, there's also something amateurish about them. I don't know whether I like them or not. What does the DIYish nature of these ads say about Canonical? About Open Source?

Monday, December 10, 2007

Does My Marketing Suck?

I've long believed that you can tell a lot about a company's fortunes, both present and future, from their marketing campaigns. Case in point: Apple. I'm far from the first person to say that their "Get a Mac" ad campaign is brilliant - both brilliantly conceived and brilliantly executed. But I may be the first person to have recommended to their friends and family that they buy Apple shares upon seeing that campaign. Why? Not just 'cuz it's brilliant. Also because of what it says about their growth strategy - namely through increased desktop and laptop sales at the expense of Microsoft. And anyone familiar with the desktop/laptop market knows that, about a year ago when Apple launched those ads, they had less desktop share than Linux, which is really saying something. So, there was a lot of room to grow - no place but up, really. After posting the biggest gains of any computer maker in the last couple quarters, Apple is up to somewhere between 3 and 5%, depending on which analyst report you look at, and their stock is up 125% since Oct. 2006.

Is it all because of the ad campaign? No, it also has to do with the blockbuster success of the high-volume iPod and iPhone, and the brand affinity these products have engendered. The Get a Mac campaign turned this reinvigorated awareness of and affinity towards the Apple brand into Mac sales - I'll say it again - Brilliant!

Similarly, sometimes you can tell when a company is getting a little bit desperate by their campaigns. Take the following banner ad, for example, which I saw running on SourceForge recently.



I don't know if / why your monitoring software sucks, Hyperic, but I'll be happy to tell you why I think this ad does.

  1. It's WAY too frontal in at least three ways.
    1. Many sysadmins that see this ad will have had some part in selecting/implementing their current monitoring software, and so they will have some degree of personal and professional investment in it. This makes them unlikely to freely admit that it "sucks," let alone take the time to tell you why just so they can get added to some spam list. It reminds me of how a marketing professor in B-School explained why the "...for Dummies" book series was successful but the me-too "...for Complete Idiots" series flopped. People's egos can handle admitting that they're a Dummy on some subjects, his logic went, but to call oneself a Complete Idiot, that's just too harsh.
    2. In my 10+ years experience in b2b technology marketing, rarely have I seen it pay off to go around poking your competition in the eye. As much as I dislike buzzwords, coopetition is the dominant structural environment in IT, and this kind of junk marketing limits a company's opportunity to collaborate with competitors in areas where it makes sense.
    3. Most enterprise customers really don't like to see vendors bashing each other - it comes off as tres gauche.
  2. It's trying too hard to be cool, though I'll admit the magic 8 ball gimmick is clever. Still, this ad's forced effort to connect with sysadmins reminds me a little of the Splunk t-shirts from a while back that read "Taking the SH out of IT" - which were good for a cheap laugh, but ultimately were too crude to be considered professional.
  3. The voice seems off - to the casual glance, which is all most banner ads get, it looks like they're asking me to tell them why Hyperic sucks (we can start with the Miami Dolphinesque color scheme).
  4. It's too obviously self-serving. Why not just save everyone the time and trouble, Hyperic, and tell us why you think everyone else's monitoring software sucks?
  5. Lastly, and related to the previous, what this ad ultimately is telling the market is that Hyperic thinks its monitoring software is far superior to everyone else's, and that, further, the company just can't understand why the rest of the world doesn't see it the same way. Translation: the rest of the world doesn't see it the same way. Which is why this ad comes off as desperate.
Add to Technorati Favorites