News Stay informed about the latest enterprise technology news and product updates.

IBM Blue Cloud: More than thin air

IBM revealed more details of its Blue Cloud computing initiative at Pulse 2009 in Las Vegas.

John Willis
IBM's recent Pulse 2009 convention in Las Vegas reminded me of one of my favorite books by Jon Krakauer: Into Thin Air. It's about an adventurer journalist who is sent on a Mount Everest expedition to expose the commodization and commercialization of Mount Everest expeditions. Krakauer's original premise is that the days of Edmund Hillary and Tenzing Norgay are stories of old and no longer as relevant as they once were. Nowadays any doctor or dentist with $75,000 and a good backpack can get to the summit via a guided tour. At least before his trip to the peak, that was Krakauer's assumption. A few broken ribs later and one of the deadliest Mount Everest expeditions ever, Krakauer found an unalienable truth: Mount Everest is a force of nature that should never be taken for granted.

Prior to my attendance at this year's IBM's Pulse event, I had become somewhat numb to IBM's cloud efforts. In some cases, I was frustrated with all the announcements of IBM's Blue Cloud. Everywhere you turn there were articles on "IBM's Blue Cloud this" and "Blue Cloud that." The simple fact was that no one could explain where you could find this "Blue Cloud" or exactly what it was. Now, a week after attending the event and with no broken ribs, I can tell you that like Mount Everest, IBM remains a force of industry and should never be taken for granted in any technology space, particularly cloud computing.

Some history

About a year ago, IBM created a working group of IBM fellows and significant thought leaders from the SOA, ITSM, and provisioning groups to determine what it should do on the cloud. This group talked with different product groups as well as some of IBM's largest customers. Over the past year, a lot of the work that has been done at IBM, culminating with the recent Pulse event announcements, can be described with one of the industry standard cloud computing taxonomies known as SPI. The SPI (or SaaS, PaaS and IaaS) model is a common taxonomy for describing cloud computing. Here's a breakdown of IBM's cloud computing offerings in these three areas.

Software as a Serivce

Lotus Live Facebook for the enterprise. Although IBM was not promoting Lotus Live as much at Pulse this year, it did show a solid checkmark with its coverage of the SaaS space. Lotus Live is what I would call a Facebook for the enterprise. I have played around with some of the features of Lotus Live with its predecessor called Blue House. Lotus Live includes Notes, Web Conferencing, Instant Messaging, and Reporting and Charting.

Platform as a Service

Have you received the memo? In this area, everyone at IBM has received the memo. It says that all product groups need to figure out how to play with the cloud. An IBM product manager described its primary Platform as a Service (PaaS) play as a block and tackle for middleware in the upcoming cloud computing revolution. IBM middleware is and has been a strong component of enterprises' participation in the Web 2.0 revolution. They want to make sure they have the right stuff for this new cloud computing switch.

As I talked to different product groups within IBM during the conference, it was clear that all of them were gearing up for the cloud. In fact, most had already produced significant cloud infrastructure components. The Rational Software Architect product has a new feature where cloud "server" topologies can be created on a workflow screen and then saved as configuration (i.e., XML) data to be used later to create clouds. Tivoli Provisioning Manager has been retooled in a number of ways to support cloud computing architectures. TPM 7.1 can provision most of the common virtualization platforms and already has working prototypes for provisioning public clouds like Amazon Web Services. The WebSphere products have been tooling for the cloud revolution for some time now with their Virtual Enterprise releases.

Middleware on steroids. Another significant sign that IBM's "Got cloud" is its new offering of middleware products such as DB2, WebSphere, and Lotus offered as Amazon Machine Images (AMI'). With these new deliverables, a developer can fire up Amazon cloud instances and start working immediately with IBM's enterprise middleware software products. IBM also announced a pricing model for product use of the IBM middleware by publishing its PVU (Processor Value Units) models for Amazon instances. An enterprise with IBM middleware licenses in place can move production resources over to Amazon's clouds immediately by using their current PVU based license model.

It's all about services, stupid. Probably the most promising offering in the PaaS space for IBM is its new TSAM offering (Tivoli Service Automation Manager). TSAM is enterprise-cloud-on-a-stick. Supposedly TSAM was jointly developed with JP Morgan Chase as a self service portal for managing internal clouds. One of the most difficult aspects of managing private clouds and public clouds is the end to end management of resources. Since most clouds are based on virtual instances, the images used to deploy cloud instances become a huge source of management chaos.

With cloud computing, it is very easy for an individual or group to create virtual servers on the fly. Also image sprawl and image identification become very real management problems. The idea that a rogue contractor or employee could take companies secret sauce (i.e., an image) and sneak it out on a USB stick is a frightening thought. Cataloging and location information of services can become impossible to manage even within the small enterprise. So it's no surprise that IBM would apply tried-and-true solutions such as service management to solve this problem. IBM has been able to plug in some of its strongest ESM products into TSAM. It includes service management products, formally MRO, from the Maximo acquisition. TSAM also includes core provisioning solutions from the Think Dynamics acquisition in 2003.

At its core, TSAM is a service management tool that provides service catalogs, a configuration management database (CMDB), incident, and change management to cloud provisioning. In TSAM all services are defined in the service catalog. Each service can have a template design for how it is to be deployed in the cloud. An order entry system, for example, might be made up of multiple systems including DB2, IBM HTTP Server, and WebSphere e-commerce. If a business owner selects an option to deploy this service to the cloud, a standard change management process can be applied for approval processing, and then the automated provisioning can be invoked. TSAM can also include the IBM Tivoli Monitoring product as a plugin to add true autonomics to the deployed cloud.

Some of the TSAM product managers and developers told me that they already have working customers with integration for Citrix XenServer as well as other third-party CMDB solutions (e.g., BMC's Atrium). Since TSAM has a robust plug in architecture it is easy to integrate with virtually any third-party service management or cloud offering. Can you use Amazon Web Services provisioning with TSAM? Well, IBM was not officially saying this; but company representatives told me off the record that they have a working prototype and to stay tuned. Having worked with IBM Tivoli products for many years, I can tell you TSAM is one of the most exciting products I have seen come out of the Tivoli software group in more than 15 years.

Infrastructure as a Service

After leaving Pulse, I was a bit disappointed with IBM's Infrastructure as a Service (IaaS) offerings. There were some bright spots, though, with IBM announcing a new Storage as a Service offering called Tivoli Storage as a Service. I was hoping for a stronger IaaS story than just storage as a service. I was hoping for something in the neighborhood of an Amazon Web Services public cloud offering or more like a 3Tera private cloud offering. IBM had some examples of private cloud customer success stories; most of them, however, were just great provisioning and server consolidation virtualization examples. IBM also had some public hybrid cloud stories with what they were calling "cloud spillage."

IBM described "cloud spillage" as provisioning additional resources on a public cloud when a private cloud has met its capacity. But in reality, most of these were really just bare-metal (nonvirtualized) provisioning stories of IBM server on-demand resources (IBM Blade Server Hosting). Even after I pinned down one of the IBM senior executives about IBM's IaaS public cloud, he admitted that today their public offering is just IBM Computing on Demand, and it's not even a virtualized technology. With, say, Amazon's public cloud, a server can be provisioned from anywhere in about 10 minutes; in contrast, an IBM on-demand server has to be physically requested and might take a whole day to be fully provisioned. This type of physical provisioning would not be in the category of what I call "autonomics" (i.e., where a server automatically detects the need for additioanl resources and, without human intervention, provisions more servers).

I did get the sense that a future storm is brewing in IBM's IaaS, although none of the IBM IaaS folk would spill their guts -- even after a few Heinekens and Coronas. Another bright spot for IBM and IaaS cloud computing was the announcement with Juniper Networks on the formation of hybrid, or rolling, clouds. But for now, this space is a little early in the works.

IBM GTS offers cloud consulting services

IBM Global Technology Services (GTS) will help enterprises determine which cloud formations are best for them. This follows the recent announcements by Capgemini and CSC stating that they are setting up cloud consulting practices. IBM hopes that its customers will come to IBM first when deciding to use the cloud.

One of the most telling tales of the IBM cloud computing stories came on my last night at the Pulse conference during a dinner meeting with members of the IBM team. One of the IBM product managers told me that IBM/Tivoli has been trying to get the entire IBM product portfolio precuts integrated for over the past 10 years to no avail. I can remember at least two or three IBM wide initiatives where IBM has tried to create a seamless, not a smoke and mirrors, interoperability story between the IBM products. Most of these previous efforts have failed. However, somehow with the excitement and I guess promise of cloud computing, IBM might finally pull it off.

In the end, I guess you can say I went on an expedition to Mount Everest (IBM) and found out something that had always been pretty darn obvious to me: IBM is a force that should never be taken for granted -- at least not on something as important as cloud computing.

John Willis has worked in the IT management industry for 30 years. He started as a tape operator on an IBM mainframe while working for his high school computer club, and began his professional career at Exxon as an IT infrastructure analyst. He is the founder of four successful startups over the past 20 years and is currently the CEO of his self-funded Zabovo Corp. Willis is known internationally for his IT Management and Cloud blog. He also has two podcast series on clouds called "Cloud Cafe" and "Cloud Droplets." Willis is also the co-host of Redmonk's "IT Management Guys" podcast series.

Dig Deeper on Cloud infrastructure monitoring

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.