CA Community






April 2010 - Posts

Cloud Expo: Some big ideas, big investments, and even a customer or two

Published: April 29 2010, 09:25 AM | no comments
by Jay Fry

Cloud computing events are happening fast and furiously now. SYS-CON's Cloud Expo was last week in New York. Interop has a cloud computing track running in Las Vegas this week with a similar set of content to the recent Cloud Connect Santa Clara show. You really could spend all your time at cloud computing events these days.


I've been picking and choosing which events to sample to get a good feel for what's going on in the market (a little bit of a science in and of itself). I won't be at Interop but will be interested to hear how different the discussion is from what went on at Cloud Connect Santa Clara. However, I did hit the Cloud Expo show, and thought I'd share a few thoughts:


· I wasn't shocked by the preponderance of vendor pitches. Why? First, the way SYS-CON is able to run the show is by offering speaking slots to its sponsors. The content logically follows suit. Think of it as a good way to sample what (some of) the players are saying and what they are focusing their investments on. If you know what you're getting on your way into the door, you can evaluate the show for what it is, rather than for what you hope it could be. One note of caution, though: several big players (like Cisco) weren't represented. Others -- like Oracle, shifting from cloud skepticism to full-on cloud cheerleader, and even smaller players -- took full advantage of that.


· Investment in cloud computing continues to grow, with signs of actual customers. These were two important data points that I took away from the event: the level of investment by the many players represented is strong. And, if you hold a cloud event in New York, some actual customers will show up. The general consensus seemed to be that the end users made up no more than 20% of the attendees based upon the number of folks raising hands in the various sessions. That's certainly an uptick from the Santa Clara version of this same event. On the show floor, the 3Tera team thought it started slowly, but liked the customer traffic; the Nimsoft folks also had good things to say (they looked pretty busy from what I saw - a good sign).


· SYS-CON got a lot of flak for their fall Santa Clara show - and listened to some of it. I was one of the folks that had a pretty negative view of the previous conference. It may just be that the market is maturing, but I was pleasantly surprised that Jeremy Geelan and crew helped pushed the conversations in New York forward, as opposed to getting stuck on the "So how do you define cloud computing?" question repeatedly. That's good to see. The next real improvement, though, will come when SYS-CON can convince the speakers to focus more on customer issues and less on sales pitches (see first bullet). There were certainly a few that did that, but not nearly enough. Hence, more grumbling about sales pitches.


· Big visions, modest means. One of telltale signs that we've got an interesting market developing - and aggressive innovation underway - around cloud computing is the healthy number of start-ups. One of those start-ups, Abiquo, decided to use a bit of its VC money to fund a stealth-until-the-day-of-the-show platinum sponsorship, giving them a big stage for their equally big vision. Their CEO painted a pretty interesting description of a "resource cloud" and a "top-down," "high-level view of the future of cloud management." They also talked about being focused on tactics for helping customers. Having been with vision-heavy Cassatt, I'm probably hypersensitive to the challenges I expect Abiquo to have trying to be simultaneously strategic & tactical (oh, and both open source & commercial, too). You can like their story, but remain skeptical that they can pull it off. Same with interesting ideas from folks like AppZero and CloudSwitch. But this wouldn't be Silicon Valley if you didn't try, right? Success doesn't necessarily mean a huge customer base and an IPO anymore, after all.


· When your data center gets hung up in customs... If you thought Abiquo's sponsorship was a last-minute thing, you should have seen Microsoft trying to get their containerized data center demo onto the show floor - and then get it working. Maybe it was intended to show off Tradeshow Booth as a Service (on-demand, you see) to add to the Microsoft as a Service mantra they were talking about. It's actually pretty impressive: no one else but Microsoft gets to create a *aaS acronym from their own name and not get major grief. Something for us all to aspire to, I guess.


· A volcano puts the spotlight on the impact of Europe on the cloud computing market. The eruption of Iceland's Eyjafjallajökull volcano and the resulting disruption of air travel across Europe made it very hard for many of the presenters scheduled for Cloud Expo to make it across the Atlantic to speak. But it also made another point: there are a bunch of European companies and personalities playing key roles in the cloud market. (See my earlier "scientists v. cowboys" post on cloud computing in Europe for some thoughts on this.) SYS-CON had pre-conference sessions with missing European presenters do some presentations using Skype. Others, like Mark Rivington from Nimsoft, sent U.S.-based replacements. Only a handful, however, showed true commitment: the aforementioned Abiquo folks were some of those. As I heard it, CEO Pete Malcom's trip to New York started with a ferry from London to France and involved a VERY long drive to Madrid, before finally finding a plane able to head to the States. "I asked how he planned to get back," noted Carl Brooks of SearchCloudComputing on Twitter. "He just laughed." Diego Parrilla (known as @nubeblog on Twitter, specializing in Spanish-language cloud tweets) figures their true-to-life travel saga is worthy of a book. Describing the volcanic cloud of ash causing everyone's troubles, he tied it to one of the hot topics of the day: "Now that was really cloud vendor lock-in!"


· The benefits of passion and a singular vision about cloud computing. I talked about the excitement that some of the start-ups showed for their ideas. That same excitement was obvious from Amazon, but for different reasons. I've heard folks discuss Amazon's culture of secrecy from George Reese (@georgereese, CTO of enStratus) and others, but that didn't come through. Instead I heard a very singular focus and world view. This was the first time I'd heard Steve Riley (@steveriley) of Amazon and he didn't actually dismiss private clouds, but at the Cloud Camp session did a pretty good job of making it obvious he didn't see a lot of use for them when there are public cloud alternatives (like, say, theirs). But the way I figure it, when your business plan is very precise, you can afford to be very passionate. And he is. And entertaining. Who else finishes up his of discussion of why private clouds aren't really clouds by saying, "I can put a goat on my front lawn, but that doesn't make it a lawnmower"?


· And, yes, the private cloud debate goes on... The Cloud Club Unpanel was debating that topic (including Steve Riley, as just mentioned), but most of the rest of the presentations took the private cloud as a given and moved on. As they should, at this point.


· For CA, a place for the new team begin to come together. For us, this event was also a good opportunity to bring the newest members of the CA Cloud Business Line team together in one place. The recent acquisitions of 3Tera and Nimsoft were the ones on display, and it was an excellent opportunity to get to know each other, talk about our different areas of focus and expertise, and make connections that are always best made face-to-face. We'll be talking more about our latest and greatest (cloud and otherwise) at CA World in a few weeks, but there are obviously some great ways to build on the existing efforts of 3Tera and Nimsoft under the CA umbrella, especially for the benefit of MSPs and other cloud service providers. Stay tuned.


So, all in all, a useful event. Perfect? No.

I was glad to connect with a more East Coast-focused set of folks working on cloud computing. One of those people, Scott Sanchez at Unisys, did a write-up of his own that's also worth a read.

At the very least, it was a good look around the industry as I help our various teams prepare for the big happenings at CA World. Hope to see you there.

 

This piece is cross-posted at my Data Center Dialog Blog.

 

Share this post:  

 

By: Jay Fry
Jay Fry is vice president of marketing, Cloud Computing, at CA Technologies. He has over 20 years of experience in marketing and management for innovative enterprise software companies. Prior to CA, Jay was vice president of marketing at cloud computing start-up Cassatt and founded the marketing department...
Read More..

Using Service and Metric Data to Manage in the Cloud

Published: April 22 2010, 01:41 PM | 2 Comment(s)
by Andrea Westerinen

Operating effectively in the "cloud" can be characterized by many things, including elasticity and agility, self-provisioning, virtualization and "pay as you go" operation. This  requires:

  • Knowledge of the services that you need
  • Knowledge regarding how those services are executing

In other words, you need to be able to query for specific services by type (what functionality is provided), and then decompose these high-level instances into the specific entities and resources that are deployed.  Lastly, you need to acquire and "roll up" metrics from the specific entities to understand and optimize the operation of the services. 

For example, if my online product catalog "service" is not functioning adequately, is it because ...

  • 1. The underlying platform and/or storage services are not functioning properly.
  • 2. The interfaces between the platform and storage services are experiencing problems.
  • 3. The code running on the platforms, using the storage, is incorrect.

Let's look briefly at how CA's Unified Service Model (USM) can help to find, decompose and measure services. Service is a core semantic of USM, defined as a business or IT offering that delivers value to an organization, and separates the consumer of the value, from the offering's implementation and operation (the latter being assumed by the "provider"). So, there is a clear provider-consumer aspect.  Information about a service in USM includes its availability, type (key functionality provided) and lifecycle state. So, basic existence and state are captured.  But this is not enough.  USM also defines the semantics to relate a service to its users, components, dependencies, impacting incidents and change orders, and more. 

Being able to instantiate various relationship semantics and being required to do so, however, are very different. USM allows products to instantiate what is "known". It places no structural (i.e., relationship) constraints on describing the physical or virtual systems that host a service or its components. This allows operation and management in an environment where little or nothing is known about the underlying platforms and infrastructure. So, USM provides the flexibility to fully define physical and virtual systems, and the relationships between these ... or not. There is no mandate (for example, via cardinality constraints) that you "must" instantiate a host system.  At a later time, an entity's properties can be matched with the instance data from other management products, to obtain a complete and accurate picture of the Service, its components, their "hosting" systems, and on and on.

Services, in this regard, are clear cut, but metrics can be problematic because there are a wide variety, each targeted for use with different resources and abstractions. Further complicating matters, metrics span a broad spectrum of information, such as utilization data, response/repair time details, cost information, and compliance and security data (for example, number of invalid login attempts for the last 5 mins at an application).  In addition, the available metrics and their data collection frequencies can vary at an individual, instance level!

To support these "ugly" realities, metrics in USM are accessed via custom operations, not defined as specific properties within a static type definition. Two operations currently defined in the USM Schema are GetAvailableMetrics and GetMetricsValues.

  • GetAvailableMetrics - Returns a list of metrics for a specific instance. The list is an aggregation of the metric data from all products that have a "view" on that instance (i.e., that have published an instance that was correlated).
  • GetMetricsValues - Returns the values of one or more requested metrics, for a specific instance. The returned data can be either live data or historical/time-stamped.

Using custom operations allows the metric data to vary at an instance level and over time.  Also, it allows the expansion of metric-related capabilities and functions in subsequent releases of the Schema. However, with all this flexibility, you are probably saying "where do I start?"  To address this, USM includes a dictionary of the metrics available from CA products accessed using the Catalyst platform. 

CA also will be working with customers, and the standards and research communities to define the key metrics for cloud/service management and optimization. This is just the "tip of the iceberg" on metrics.  There is much work to do!

What are your thoughts on critical service types and metrics?  We'd love to hear from you.

 

 

Share this post:  

 

By: Andrea Westerinen
Andrea Westerinen is Vice President of Software Engineering, and an evangelist for semantic technologies, working in CA’s Office of the CTO. She is responsible for the development and oversight of CA’s Unified Service Model (USM), and for the design of innovative infrastructure to better align business...
Read More..

The Cloud leans back – an iPad epiphany

Published: April 20 2010, 08:32 PM | no comments
by CA Community

Given the giant cloud of volcanic dust currently passing over Europe I could have also titled this blog "The cloud strikes back", but that is not the topic of this "the cloud changes everything" blog.  It also has nothing to do with my earlier topics on Lean IT and Lean manufacturing. Lean Back and Lean Forward were the guiding principles for developers of Interactive Television.


 

The idea being that TV is typically enjoyed leaning back, while computers are typically used leaning forward. Of course we all know that interactive television so far has been as successful as the NEXT computer (not). In fact, in the last 5 years we moved to a situation where more and more entertainment content is consumed leaning forward . If you visit our home on a typical Friday night the big ass TV may be on, but meanwhile all family members are hammering away on computers. Mom on the desktop, the kids fighting over who gets the macbook versus the laptop and Dad on his company supplied artifact.

Now only a few years ago, you assumed that any person using a computer was probably working. (back then only dad would be using his computer on a Friday night). One of the first I saw point out this blurring of work, home and play - as illustrated below -  was IT visionair Peter Hinssen (t) , in one of his hilarious but at the same time highly educational sessions on IT strategy.



However, something felt wrong about this. Maybe it is a generation thing, but if the above is your life, then how much of a life is it?

Luckily help has arrived, but not from interactive TVs.  If a family of 4 can barely agree what TV channel is to provide the background noise to their computing activities, how would they agree on one joint interactive activity. This new reality was vividly described by Robin Bloor , one of the first industry analyst and author of havemacwillblog.com.  In how  "The iPad Will Replace The Laptop" he describes numerous use cases of how "lean forward" is no longer needed to have a good and/or productive time online. If you're a person dying to get an iPad (like most of us in Europe) but also if you interested in the future of TV or indeed ... in the future of the cloud ... then this is a must read story.

Important to realize is that this is NOT about switching devices, but about a  fundamentally different way to interact with content and functionality (a.k.a. with the cloud).  People are interested in taking a ride, not in the art of motor cycle management or ownership. And that is exactly what the cloud brings. In fact, Google referenced the same phenomena when launching their latest incarnation of Google Docs. They stated that about as many people should need an office suite as need photoshop, autocad or similar. Most people read or at best annotate, only very few create. 

In this light, the decision of the Library of Congress to preserve the first billion tweets for prosperity (including interesting twitbits "like what did Dave or Joe or Bob  have for Lunch on Monday Aprill 11th" may not be so stupid after all. The "we are all authors" movement may turn out to have been an one-time phenomena caused by temporary ubiquity of keyboards. Let's face it, writing 2500 tweets for 7 followers is not a fulfilling, economic or even sensible activity, let alone reading them (if you don't believe me have a try). Similarly we may realize that the volume of email we all complain about , is because we all spend too much time typing and not enough time reading. 

I think we can all agree Apple already changed the computer industry several times and is revolutionizing the music industry. With the iPad they will now be changing the TV industry (Apple TV was not a failure, it is just infrastructure for the iPad)  and by leaving out a keyboard they now also created the ultimate work, home and play "cloud consumption experience".  




 

This may mean that we are going back to an time where people using PCs are not just the ones working .  They are likely to be the ones creating cloud stuff: content creators and functionality developers (a.k.a . IT folks).  All other people - working, relaxing  and playing - will be doing so on cloud consumption devices.


Case in point, from a tweet by Cap Gemini's CTO Ron Tolido: "Need to run a country? We have an app for that" that points to Norway's prime minister running the country from his iPad while stranded due to the aforementioned volcanic cloud.  So lean back (or lean forward - depending on your job area) and enjoy the ride.


 







PS Some of you may point out that for that for the above to work, it would be a good idea if Google Docs actually worked on the ipod . In typically IT fashion we would call this an "implementation detail", but feel free to point it out to Google, preferably before I get a chance to pick up my iPad from Europe.

Note: I could not find the original author or copyright holder of the Modern Life cartoon, any pointers appreciated.

Share this post:  

 

By: CA Community
CA Community is the blog manager’s account used to post general updates and news items.
Read More..

More Posts