Who’s On Phirst

Evelyn Hubbert of Forrester Research just came out with a new report, “Low Hanging Fruit That Service Operations Teams Should Consider Now”, which details some of the IT tools that organizations can use in the short term to increase efficiency and save money. I had the chance to have lunch with Evelyn this summer and get some of her thoughts on the IT tools landscape, and boy, does she know this space well.

Here are some of the nuggets from the report that I think everyone in IT ops should be thinking about:

• Approximately 75% of the IT budget is spent on simply maintaining existing IT operations, and IT organizations must evaluate automation solutions.
• Leverage existing tools to their fullest, but also look for complementary tools that can support your focus area and bring immediate benefits of improving efficiency, reducing risk, and supporting end-to-end business services and at the same reduce cost.
• In 2009, the general IT battle cry will be to do more — or at least as much — with less and that most IT organizations will be able to keep their IT staff constant at best.

In evaluating tools, IT organizations are going to have to look at fast implementation times, and fast returns on their investment. Tools will need to be able to show how they can either cut costs directly or allow current staffing levels to scale to more support more assets. Or both. The time of soft ROI is over. Paybacks should be measured in months, not quarters.

Click here to read the full report.

In Untagged 
Comment (0) Read More...

Check out this new report on Phurnace, the cloud, and automation by Julie Craig of EMA (Enterprise Management Associates).

A few of the key points made in the brief include:

• The key to Cloud success is going to be your management tooling. Doesn’t matter how fast you can spin up images if it still takes you three weeks to deploy your application.
• Infrastructure independence is key: your strategy for the Cloud has to take into account bridging the gap from the Cloud to on-premises infrastructure.
• We need more, and more sophisticated, automation.

In Untagged 
Comment (0) Read More...

I and a few other folks from Phurnace attended the Portal Excellence Conference in San Diego this week. Some quick thoughts from the show:

• Insurance and Banking seemed to be the biggest verticals represented at the show.
• There were quite a few folks there that weren’t current Portal customers, but were considering migrating from another platform to WebSphere Portal. The most common reason they gave was nervousness about all the recent consolidation in the Portal space (i.e., Oracle acquiring BEA, OpenText acquiring Vignette, etc.)
• More talk about WCM (IBM Web Content Management) than in prior years. WCM seems to be catching on with the WAS Portal customer space.
• IBM is an excellent host at these events. Thanks guys! We had a good time.

In WebSphere Portal
Comment (0) Read More...

Yesterday, the leading tech analyst, The 451 Group, published a report on Phurnace. They talk quite a bit about how Phurnace is starting to look like a “cloud foundry”. It is a great report that gives their insights after they were briefed on some of our upcoming product enhancements (Phurnace 4.0 and cloud targeted products).

One of the most interesting lines in the report is this: “It is not yet clear how clouds will be used in the enterprise, but it is likely they will underpin J2EE applications. If that does turn out to be true, the sheer scale and speed of elastic cloud deployments will preclude hand-coded provisioning. Something like Phurnace Deliver will be required.”

Please check out the full report here.

In Cloud Computing
Comment (0) Read More...

I love talking to customers. It makes product planning so much easier if all you do is listen to them and give them what they want. And from the looks of things, most all of our large WebSphere customers are planning to or have already moved to WebSphere Virtual Enterprise (VE).

Quick synopsis of VE (or the product-formerly-known-as OO or XD; IBM has played around with the name more than Prince). Virtual Enterprise allows WebSphere clusters to dynamically allocate resources to applications based on a series of policies, either for service levels or health statistics. The VE on-demand router can gather utilization information and then implement topology or prioritization changes. It’s pretty cool.

There are some challenges though. Just like anything in WebSphere (or middleware in general), it’s a bear to configure and maintain. It’s critical to implement the policies correctly and unlike static configuration where it’s often immediately evident if something isn’t right, with policy based configuration everything can look perfectly fine, but under load the environment might behave quite a bit differently than anticipated -- if the policies aren’t set up exactly right.

About 80% of our Fortune 500 customers that use WebSphere are either currently using VE or have concrete plans to move to it in the next six to nine months. That’s pretty compelling. That’s also why we have extended our products to support VE. With our VE support we help insure that our customer’s environments are set up exactly the way they want them and eliminate a lot of the complexity and heart burn that comes with the added functionality. So, rest assured: go ahead and move to VE, we have your back.

 

In WebSphere VE
Comment (0) Read More...

Alright, I keep getting asked this so here are my thoughts, in bullet form. Take them for what you will:

• First, this is what FastScale provides, in a nutshell: on demand, they would examine an application and strip out all the superfluous fluff in it to make it streamlined and easier to manage/deploy, etc. Cool stuff. But, no, not what Phurnace does.
• This is a good move for EMC. It adds to their system management suite, and FastScale provides them a compelling differentiator.
• M & A in this space is heating up. More deals are happening with better valuations. The slope is now up and to the right.
• The companies that are surviving the downturn are well positioned for strong performance moving forward.
• Notice how we are hearing the word application more and more? (application deployment, application virtualization, application management, etc.). Also notice how we are hearing the word server less and less? That’s not a coincidence. I.T. is making the correct migration from thinking about servers and provisioning them to applications and enabling them. This is good for the industry and to help align business goals with I.T. goals.

In Untagged 
Comment (0) Read More...

We are in the middle of a two week industry analyst tour right now and we completed some really good briefings last week.  Lots of discussion on the need to provide rapid ROI, the need to show immediate cost savings to IT departments, how to leverage the desire in companies to use even more automation and of course, the Cloud.  Phurnace is in a very good place related to the hottest trends in IT.  Automation, a shift to an application centric mind-set in IT, cost reduction, and cloud computing.  The analysts see this and almost before we got to slide 2 in our presentation asked the question, “what do you think of VMWare’s acquisition of SpringSource?  Is the bubble back?”  First, no, the bubble isn’t back.  While $420 million is a great price for SpringSource, this is NOT like the days of $1.2 billion for Toothpaste.com or such silly things.   Those days are gone forever.  SpringSource has a huge following, real revenue and will add to VMWare’s product portfolio in a good way.   I see the SpringSource acquisition as good for everyone.  For them, for VMWare, for the market, for the momentum of virtualization and cloud computing.

We took the opportunity with the analysts that we met to talk about the future of app deployment in the cloud and where Phurnace fits today, but even more exciting, where we will fit in the near future.  We got great feedback and our ideas were validated.  We have some exciting plans for additional products targeted at the cloud.  Amazon Web Services first, other public clouds after that.

We were continually asked, “why aren’t you partnering with VMWare?”  Actually, that makes sense.  Our software deploys applications into physical, virtual or cloud environments.  VMWare would be a logical partner.  We have just been so busy with other customer requests.  I guess since our software works out of the box with VMWare, we didn’t really see a need to call them and bother them.  I bet they have their hands full right now with SpringSource. 

In Virtualization, Cloud Computing, Amazon Web Services
Comment (0) Read More...

At the beginning of March this year, I got to spend a couple of days at the BMC Lexington campus prototyping our BladeLogic integration with some of their product, architecture, and development folks. Other than the four feet of snow, it was a great experience meeting the guys there as well as seeing what Phurnace can do with a product like BladeLogic. Together we can do lots of neat things that today just aren’t possible with either product alone. Recently one of our customers brought up an interesting use case that reminded me of one of the BladeLogic features.

In BladeLogic, you can setup heuristic rules to evaluate a task, decision, etc, and then take the appropriate action based on the rule. It’s a pretty rich interface, with full support for all first-order logic. This is pretty cool in its own right. But as they were showing the product to us, I kept thinking that if we combine this with the Phurnace data model for configurations, we can do some very amazing stuff.

For example, one of the things that scares the pants off IT operations guys is there being inadvertent changes to security or tuning settings. Both of those result in big fires and lots of unwanted attention. The problem is that sometimes they get packages to deploy that aren’t as sensitive to those requirements as they should be. Or, even worse, they get an app that HAS TO BE configured in a way that violates their policy in order to work – a hard-coded port, for example. See, the problem is the system administrators know what their environment needs for things like performance and security, and the devs know what the application needs for configuration and deployment, but the melding of those two bases of knowledge is always nasty, and often explosive.

So, watching the BMC guys take us through BladeLogic, I was thinking that this would be a great way to solve that problem. Have the SAs program in their tuning and security requirements actually in the heuristic engine, and then when Phurnace provides the data model for the configuration of the application, run a check to see if it violates any of those policies. The heuristics don’t have to be explicit – they can be a range of options, or settings, or what have you. They can be as complex or as simple as desired. But the important point here is that this check is now done programmatically, in an automated fashion, with no heartache or lost time. Good things happen when you combine good data with good process.

In BladeLogic
Comment (0) Read More...

Phurnace's VP of Products and Co-Founder, Daniel Nelson, wrote a blog posting for one our partners, Electric Cloud, earlier this week. Here is a link to their blog and here is his post.

In the software development lifecycle the results need to ultimately get out into production. The application must make it onto the server, the server needs to be configured and all of the properties, paths and settings need to be correct to get the value of the application. Those in the data center often refer this to the last mile. And it is often a real bear. Seems more like 10 miles at times. Sophisticated and robust tools like ElectricCommander have automated almost all of the steps in the process, but rely on home-grown scripts to lay the applications down on app servers. Why? It is actually a logical approach if you have no alternative. Every environment is different. Every app has different settings for WebSphere (or WebLogic or JBoss) and there is no way to anticipate those differences. Therefore – the last mile is unique to each customer and each app –the IT or dev teams write scripts. It takes skilled resources and the scripts are always in need of attention. Not anymore. STOP. That is no longer necessary. What if ElectricCommander could hand off the EAR file to a software tool that has already pre-built a model of the environment and has made all of the JDBC, JMS, and application bindings for you? What if it required NO scripting? What if the “last mile” was now automated and under the control of your build and release system? Ta Da ! It is now. Phurnace Software is an auto deployment and configuration tool that will eliminate custom scripting. And you can drive it all from within ElectricCommander. The last mile is now just a step away.

In java
Comment (0) Read More...

Ops folks hate agents. Hate them. I was in the same camp back in my data center days. The reasons are pretty obvious, right? Who wants to have to install and maintain another piece of software on every server? No one. Not only that, but agents scare the bejesus out of people running production systems. A small program that executes commands locally from a remote source is super scary no matter what. Like clowns with fangs scary. Add in the chance for that little agent to go rogue and start consuming resources or leaking memory or hording ports and scary quickly morphs into dread.

With all the revolutions going on in server infrastructure agents are also becoming more and more inconvenient. Virtualized environments and cloud computing make the servers themselves much more ephemeral than in other days past. Maintaining agents to these here-today-gone-tomorrow servers, and all the references to them, and the configurations they require, is just too time consuming and troublesome. It’s a hold over architecture from days past when “server” meant a thing with a form factor and a fan, not dynamically allocated memory and processor time.

Phurnace doesn’t play that game. We connect over the network – no agents required. That gives us tremendous flexibility to fit into the existing infrastructure architecture of our clients. Also, it allows us to configure and maintain servers regardless if they are physical, virtual, or cloudy. Ultimately our customers don’t like being told how they have to do things – they want products that fit how they want to work, not have to work the way a product demands. Being agentless is part of our flexibility that let’s our customers do that. And of course, makes us far less scary.

In Untagged 
Comment (0) Read More...