Monthly Archives: November 2011

Want to know what we do (and why)?

Our own Dave Rosenberg recently hit the press circuit, and returned with two Q&A sessions that help to explain what Nodeable does, and why it matters.

The first involves Dave’s prediction that in 2012 IT, and particularly DevOps, will seek to measure everything.

Along with the rise of cloud-to-cloud application networks, I also expect to see a great deal of interest in the monitoring and management of cloud transactions. In fact, this may be where we start to see the rubber hit the road in the DevOps world as different user groups within an organization have vested, but different interests in ensuring that data arrives quickly and consistently.

To that end, we’ll see a growth in the analysis of communication in the cloud and solutions that address everything from the “digital exhaust” associated with every layer of the stack-from each transaction to the system management aspects of virtual machines themselves.

Given Dave’s prediction, it’s perhaps not surprising to see that Nodeable is trying to solve this problem of complex system data with a clean, unified interface, as Dave notes in a separate interview:

The short version is that we struggled to make the existing tools match the new world of cloud services, which are quite fleeting. It didn’t make sense to script and automate resources that were transient, and none of the tools were designed in a multi-tenant fashion, meaning we didn’t have a breadth of deployment options.

And, the majority of the available tools are ugly and hard to use. We wanted to simplify the management tools and take advantage of the modern compute power the cloud offers….

From the Nodeable perspective, we think we can define and provide a layer in the new cloud stack – one that takes advantage of the “digital exhaust” we get from systems and adds intelligence to data. The gathering and analysis of this data leads into a systematic approach to managing these services based both on deterministic responses and predictive analytics.

Which means that Nodeable isn’t necessarily replacing other system management solutions, but rather complements them by ingesting “exhaust” from these and other systems (be it Salesforce, AWS, or Jira), thereby giving DevOps a single pane of glass through which to see relevant system events and act upon them.

Just don’t make the Nodebelly angry.

Enterprise, but not boringly so

One of the things I like most about Nodeable is that while we take systems intelligence seriously, we don’t take ourselves too seriously.  Punctuating the frequent conversations about the UI, which systems to monitor, etc. are comments about whether Jane’s Addiction’s new album is awesome (Yes, I say; no, says Dave), who will play bass for the company’s band, etc.

And then there’s the Nodebelly.

My kids love the Nodebelly.  He’s what greets you if you hit a 404 error on the Nodeable website.  He’s the guy who greets me whenever I head into the office, thanks to a whiteboard drawing of him that our wonderfully creative Mike Evans drew.  He’s what reminds me that enterprise software, just like consumer software/brands, can have a personality.

Some enterprise companies get this.  But very few.  Atlassian, 37Signals, and…not many others manage to build grown-up software that doesn’t bore you to tears with its messaging.  I suppose it’s an effort to show would-be buyers just how safe a product is, by demonstrating just how dull the company’s marketing can be.

But this seems silly.

A company’s marketing is not only a signal to prospective buyers, but it’s also one of a company’s primary recruiting tools.  The sort of people who want to join Atlassian, for example, are going to differ from those that want to join CA, which will mean a very different type of salesperson is going to end up sitting in the CIO’s office.  My money is on Atlassian to be the more interesting of the conversationalists.

All of which makes me glad to be part of Nodeable.  We’re building cool technology to manage cloud deployments, and we won’t bore our customers and partners to tears with our marketing.  After all, we wouldn’t want to make the Nodebelly even more glum than he already is.

Tagged ,

Are your management tools slowing down the cloud?

The cloud isn’t about cost, though by shifting enterprise spending from CapEx-heavy server purchases to OpEx-friendly cloud subscriptions, the cloud not only has the potential to lower costs but also to shape how they affect budgeting in highly positive ways.  No, cloud is about speed.  Speed of development.  Speed of change.

Speed of innovation.

This message is driven home in a 2011 InformationWeek report entitled “IT Is Too Darn Slow.”  InformationWeek polled a large group of IT executives, who point to SaaS and cloud as primary drivers of increased speed of IT.

This makes intuitive sense, as it allows IT to spend less time managing hardware, installing software, and other mundane tasks, and gives it more time to focus on actually driving business value.

The problem, however, is that IT management tools haven’t kept pace.  Even the best of such tools tend to be log-driven, which is by definition offers a historical view of what happened in one’s IT infrastructure, be it on-premise or cloud-based.  But given the speed at which DevOps moves, IT really needs a real-time view of what’s happening.

This is, of course, something that Nodeable does well.  Nodeable adds intelligence to systems data, making it easy to view, analyze, and manage cloud infrastructure and services…and Nodeable does this in real-time, unlike log-based systems or antiquated IT management tools.  Rather than drown DevOps in a deluge of data/updates, however, Nodeable surfaces relevant, actionable system data in a modern interface that is as easy to use as Facebook or Twitter.

The data processing on the back-end is powerful, and so is the interface used to view this real-time system state.

All of which is only going to become more important over time.  Over the next three years, Morgan Stanley Research shows (warning: PDF) a steady decline in server shipments (except to cloud providers), even as adoption of the cloud explodes.   Whether viewed in terms of migration of workloads to the cloud or enterprise adoption thereof (which the research pegs at a 50 percent penetration rate by 2014, growing at a 23 percent CAGR over the next three years), cloud adoption is real, it’s broad, and it necessitates a change in how we manage IT.

It’s not surprising to see industry dinosaurs relabeling their 20th Century solutions as “cloud management.”  But labels can be deceiving.

Real cloud management is native to the cloud, and is designed from the ground up for multi-tenancy and other aspects of cloud deployments.  It shouldn’t show isolated server logs, but rather give a holistic view into the interplay between disparate cloud systems and how such interactions are affecting performance (i.e., is AWS the problem or is it Github?  Or a combination of both?).

Services like Nodeable and Boundary, which I covered for The Register,  get this.

Even as IT hits warp speed, driven by SaaS and cloud, enterprises need to take care not to let their IT management tools get in the way.  Cloud management tools run in the cloud, deliver real-time system data, and provide a contextual view of what’s happening in the system.  In short, they make IT even faster.

Yes, your cloud management baby is ugly

As perhaps the most hyped term in computing today, “cloud” has attracted no shortage of vendors desperate to tap into the burgeoning cloud-related IT market.  In just public cloud-based storage alone, IDC expects enterprises to spend $11.7 billion by 2015, up from $3.3 billion in 2010. The stakes, in other words, are huge.

But so is the complexity, given the number of vendors competing for wallets.  Earlier this year CloudTweaks offered 15 (actually, 18) recommendations for cloud management tools.   A few months later ReadWriteWeb pointed to a couple of others, SearchCloudComputing.com highlighted others, and most recently VMware got into the act with its own cloud management tools.  The ending tally includes more than 30 different solutions, each with a different take on collecting and distributing data.

Then again, maybe not.  While RightScale is very different from Morph, which is different from HP OpenView, most of the cloud management solutions offer a somewhat standard suite of status reports (on uptime, response time, etc.), plus a dashboard.  I don’t even have to include a screen shot here for most system administrators to have a pretty good idea of what those reports and the dashboards will look like.

They’re ugly.  They’re not overly helpful.  And they’re 20th Century.

This is one reason that I joined Nodeable.  There are actually lots of interesting things we’re doing, but one that is near and dear to me is that we’re breaking the mold on traditional aggregation and display of cloud service data, with a Twitter-esque UI that should appeal to the rising generation of system administrators.

This isn’t some gimmicky way of viewing status updates.  As my friend and Nodeable CEO (aka “Chief Nodebelly”) Dave Rosenberg explains, Nodeable “adds a social layer to systems data, allowing users to collaborate through activity streams instead of traditional dashboards and reports. Ultimately, it’s a collaborative systems management tool for the cloud.”

This approach makes cloud management and data services much more intuitive.  Anyone that knows how to use Facebook or Twitter will feel right at home in Nodeable’s interface, and might even find the talkative machines (“@Opsguy HTTPS request for #port443 has exceeded 46ms”) far more interesting than former girlfriends and current flesh-and-blood neighbors.

End-users have become accustomed to new ways of communicating and getting things done outside of work.  In Nodeable’s world, ambient awareness and contextual collaboration replace traditional dashboards and reports…just as they do in our non-work time.

There is more to Nodeable than a clever way to consume and collaborate around systems data.  The company also provides management actions for a growing variety of data sources, starting with Amazon Web Services but extending well beyond to GitHub, Atlassian’s JIRA, Puppet, and more.  We’re working on clever ways to sift through the multitudinous data that cloud systems can yield, highlighting only that which a cloud developer needs to get her work done.

But I admit to being a big fan of our interface, and its possibilities for making arcane systems data comprehensible and actionable. Gathering data is a difficult but doable task.  Making it actionable and intuitive is dramatically more difficult, and Nodeable does this in a unique manner that appeals to today’s social media-savvy system administrators.  I wouldn’t expect to see yesterday’s vendors like IBM or HP taking a similar approach anytime soon.  It’s simply not in their DNA.

Those interested in getting a peek for themselves can register for the Nodeable private beta.

Tagged , ,

Welcome new Nodeable team members

This week we added not 1, not 2, but 3 new members to the Nodeable team but we’re still hiring. Right now our search is primarily in engineering and data services where wizards and goblins make sense of vast amounts of unstructured data.

Matt Asay  Matt Asay (@mjasay) is our new SVP of business development. While we have no idea what that title means, we are sure he’ll be a great addition to the team.
rohinip Rohini Pandhi (@rohinip) joins us in product marketing, as a fairly new MBA graduate. You will marvel at her Excel skills.
rohinip Tim Golen is new to our development team. He loves Jquery and all-natural beef.
Follow

Get every new post delivered to your Inbox.

Join 53 other followers

%d bloggers like this: