Who’s On Phirst

A few months ago, I went to France to help my friend, Will, work on his house. He lives in New York and travels to France once a year to put in some work. In turn for manual labor, he picks up the accommodations tab and buys us wine.

One of my travel buddies was Will’s brother-in-law, Henry. He’s a great photographer and has a killer digital SLR camera. He took lots of pictures. However, once I checked out the pictures online, I realized that the only way to truly appreciate the shots he took was to have the pictures developed and printed on photo paper and put in an album. Given that I was simply going to view and share the pictures online, wouldn’t a point-and-shoot model have worked just as well? Moreover, we were constantly making sure Henry’s camera didn’t get stepped on or stolen. I’m certainly grateful for the pictures, as I didn’t take any, but I can’t help but think that the pictures would have been just as good if he had a lower-end camera.

Robert Capps wrote an article on that same concept. Sure, vinyl LP’s have the best audio reproduction, but an MP3 is so much more convenient. In the long run, your ROI might not be as much with an MP3, but you’ll be able to see a return much faster. Sometimes, that’s just good enough for songs with a short half-life.

When it comes to Java Application Server management, there are quite a few products on the market that are simply overkill and don’t solve your true problems. The installation is painful and difficult, an army of consultants are needed for integration, and the price is outrageous. Moreover, they don’t solve your problems with configuration management and application deployment NOW. The only thing they are good at is separating you from your money.

Sometimes, an easy-to-install, easy-to-use tool is all you need. But, here’s a big secret: Phurnace Deliver isn’t just a disposable point-and-shoot camera. It’s a samurai sword tucked inside a Leatherman made out of adamantium.

Our customers see less than 100-day ROI after purchase. No other solution around, homegrown or purchased, can get close to a 100-day ROI. If you think you found one, let me know. I’d love to have a public Pepsi Challenge.

In Untagged 
Comment (0) Read More...

Like it or not, you have a heterogeneous environment. You are using a variety of OS’s, hardware architectures, databases, you name it. And, it’s proliferating. As companies continue to seek the best bargain for their IT purchases, they are more willing to use a different technology in order to stretch a dollar.

We’ve seen this shift for some years now with open source software. Companies that are in need of a database for a specific application are less likely to purchase more DB2, Oracle or SQL Server licenses and more likely to build the application using MySQL. The same can be said for Java middleware. At Phurnace, we have seen a large portion of our customers moving from WebSphere and WebLogic to open source alternatives such as JBoss or Geronimo.

Typically, these decisions are being made by internal development groups or by the companies’ external vendors. For example, we are good friends with two software vendors in Austin that have migrated their applications to support JBoss exclusively. Thus, they provide their customers with a complete stack and not just the application.

For internal development groups, the decision to build their application on an open source alternative to WebSphere or WebLogic is driven by economics and ease of use. Typically, the development group does not have the resources nor desire to use a non-free middleware platform.

The danger here is caused by lack of experience within the IT organization tasked with managing the application and its middleware platform. First comes the challenges of learning how to manage a completely new application. In turn, this is compounded by learning a completely new middleware platform. Often, determining if the cause of an application failure is due to inexperience with the middleware or the application itself is near impossible.

For this reason, we support JBoss, WebSphere and WebLogic. Our customers have the ability to manage existing middleware configurations and application deployments and can use the same tool they are familiar with to manage all of the platforms. Or, even better, this can help with their migration from one platform to the next.

Simply put, Phurnace wants you to be successful with your application server middleware choice, regardless of what you choose. Don’t worry. You won’t hurt our feelings when choosing one platform over another. We support the three most popular today and will be adding support for more.

In Untagged 
Comment (0) Read More...

At Phurnace, we take great pride in our customer service. Since day one, we have created processes and employed systems to help us reach our goal of 100% Customer Satisfaction. A world class product like Phurnace Deliver is only as good as our customers’ ability to implement our solution. To that end, we utilize Salesforce.com’s Customer Portal and provide our customers the ability to directly contact our engineers. We are so confident with our solution, that every single one of our customers has my mobile number.

To reach our goals of 100% Customer Satisfaction and sub-100 day ROI, we partner with our customers and don’t simply sell them software. Our customers drive our product road map and feature list. We have chosen that route because, simply put, it’s easy. After all, why throw darts at a wall of new features when we can simply ask our customers what they want?

To that end, I have found that two qualities in our customers leads to greater success and faster ROI: being communicative and having a desire to change.

Communicative: Simply put, we want our customers to tell us what’s they need. To help us answer any of their questions or to guide the future features, we have adopted an agile development model that allows us to shift priorities without risking deliverables. However, that model presupposes that our customers will let us know what is on their mind and what is truly important to them. If there is a request , our engineering staff is able to quickly produce hot fixes for our customers once they have been notified. It pains me considerably to hear from a customer “Oh, that’s been happening for months,” or “We ran across that in February.” Of course, once we are able to prove to our customers that we aren’t like other software companies and that we will actually fix our problems, that communication comes much more freely.

Desire to Change: Our product requires a shift in how customers think about managing their configurations and deployments. Instead of thinking linearly about the mechanism of change, we ask that our customers just think about the change itself. “Concentrate on the ‘what’, not the ‘how’” is what we like to say to them. For some of our customers’ team members this is difficult as it does require them to approach problems in a different manner. As with all change, you should only initiate change only when you have a clear path to a benefit. To this end, we continually work with our customers to assure that they are tracking to reach ROI in under 100 days. In fact, we have sent engineers to customer sites well after training to help move them closer to that ROI target, simply to make them a happy customer.

100% Customer Satisfaction and sub-100 day ROI are lofty goals. Yet, they are attainable if you partner with communicative customers with a desire to change.

In Untagged 
Comment (0) Read More...

Today I had lunch with Duane Tharp from StreamStep. Duane is a software technology leader and has had great success with NetSuite, BetweenMarkets (acquired by Inovis) and mValent (acquired by Oracle), which he cofounded with his current cofounder, Clyde Logue. StreamStep automates and optimizes all aspects of release management. But, it’s not just limited to pushing code out the door. They can also help you with server management. Basically, if you’re using an Excel spreadsheet and hacked together scripts, you need StreamStep. They just formed a partnership with BMC and are integrated with BMC BladeLogic for Application Release Management. They’ve also partnered with Splunk to allow customers to easily search StreamStep data. Sounds like you can now answer the question “What did we deploy last Thursday?” without getting into a war room. So, please check them out and let us know what you think in the comments below.

In Untagged 
Comment (0) Read More...

There is a lot of discussion in the market right now on the Spring Framework and development language because of the recent VMware acquisition of SpringSource. First, for those of you not that familiar with Spring, it is a lightweight J2EE application platform and integration model. Spring also helps reduce the complexities involved when developing J2EE applications. For example, in the J2EE EJB model, Spring only requires you to create your Domain Module using Java Beans freeing your from a lot of code writing required by J2EE. Spring has found a place in thousands of companies worldwide. In fact, we use Spring at Phurnace to help us manage the multitude of MBeans we support.

Now, if you’re going to be running your Spring application on its own, you have a very simple deployment process. However, most companies want the ease of development that Spring provides but the robustness that a full-blown Java Application Server can provide. In this case, your Spring based application has all of the same deployment challenges as a J2EE spec application has. It still needs to have the target app server configured, tweaked and set up to correctly run the app. That is where Phurnace comes in. So, yes, Spring-developed apps as well as J2EE apps are handled by our deployment automation product.

Now, a comment on the SpringSource acquisition. The VMware acquisition of SpringSource is all about the assembling of a “stack” for the quicker and easier building and deploying of applications in the cloud. Finally, the world is starting to move to a more application-centric view of IT. No longer is the data center a place where servers run infrastructure. IT is about applications. Everything is there for the support of those applications. The cloud is an environment with a very application-centric approach. However, it will not replace the data center, but instead, augment it. AND, applications will need to migrate between instances in the cloud and between images and servers on premises behind the firewall. To and from the cloud. That is a bunch of configurations tasks or custom scripts that will need to be constantly tweaked and maintained. Unless you use Phurnace. You see, Phurnace can move the applications to and from the cloud and to and from any virtual image, regardless of where it is.

VMWare made a good move by acquiring SpringSource and it will speed adoption of the cloud for development and testing. However, the story isn’t complete without Phurnace. Thank you VMware for laying the groundwork for more robust cloud usage, we are ready willing and able to make sure the applications are actually deployed correctly with Phurnace.

Also, for more information, please read this blog on Spring and JMX posted by one of our developers: http://www.phurnace.com/blog/spring-plays-well-with-jmx.html

In VMware, Spring Framework
Comment (0) Read More...

Both WebSphere Virtual Enterprise (VE) and ND 7.0 have the ability to remotely stop and start Node Agents, install WebSphere on remote servers and a whole host of handy remote activities. To perform these actions, WebSphere takes advantage of Tivoli Remote Execution and Access (RXA).

Luckily, you too can take advantage of RXA to manage your remote servers. Below, I’ll show you how to create a very simple Java class that will connect to your local windows machine and perform a directory listing.

First, you will need to create a new Eclipse project and add the following JAR to your Build Path as an external library: ${WAS_INSTALL_DIR}/plugins/com.ibm.ws.prereq.rxa.jar.

Second, create a new class and add the following main method.

      public static void main(String[] args) throws Exception {

            RemoteAccess ra = new LocalWindowsProtocol();
            ra.beginSession();
      }

At this point, you can really take advantage of the RemoteAccess object to gather information about your Local Windows host, or to run commands, manage processes, you name it. For example, the following lines added to the main method will perform a directory listing on your Local Windows host:

            ProgramOutput programOutput = ra.run("dir");
            String stdOut = programOutput.getStdout();

Of course, you might want to get STDERR while you’re at it…

Finally, RXA is not just limited to a Local Windows host. Here is an example on how to connect to a Remote SSH host:

            String username = new String("root");
            String strPassword = new String("p4ssw0rd");
            byte[] bPassword = strPassword.getBytes();
            ra = new SSHProtocol(username, bPassword);
            ra.setHostname("myhost.domain.com");
            ra.beginSession();

If you are going to be running a series of identical commands on a bunch of different servers, all with different OS types, I would recommend using the Factory design pattern to generate a RemoteAccess object of the correct Protocol type. Below is a list of the Protocol objects. As you can see, you have a very broad assortment.

AS400Protocol
LocalAS400Protocol
LocalUNIXProtocol
LocalWindowsProtocol
REXECProtocol
RSHProtocol
SSHProtocol
UNIXProtocol
WindowsProtocol

In WebSphere VE, WebSphere 7.0, Tivoli Remote Execution and Access
Comment (0) Read More...

This week, Phurnace announced our support for WebSphere Application Server for z/OS. You can read more about it in our Press Release.

To be honest though, we have always worked on WebSphere for z/OS. However, we were unable to give it the official Phurnace stamp of approval until we ran it through our testing battery. As you know, z/OS is IBM’s mainframe operating system. Thus, for an ISV like Phurnace, procuring a z Series machine and the skills to manage it is a bit outside our skill set and budget. Not to mention the raised floor room, cooling pipes and tanks of Halon (not really, I think they finally banned that stuff).

Thankfully, we were able to secure a z/OS virtualized machine from the IBM Innovation Center in Dallas. The support we received from the team in Dallas was phenomenal. Patient and helpful, they provided us with a turnkey z/OS image with WebSphere ND installed and ready to test against. Moreover, the entire process was at no charge because we were validating that Phurnace Deliver™ would work with z/OS and WebSphere for z/OS.

We have access to the z Series system and z/OS for demonstrations. Please contact us if you would like to see it - sales@phurnace.com.

In zOS, WebSphere
Comment (0) Read More...

Last week the Phurnace team visited Las Vegas to attend IBM’s Impact 2009. We presented a “Bird’s of a Feather” talk and spent some great time with customers (current and future) at our booth and at dinners.

While at the conference, the most often asked questions were about WebSphere 7.0. It seems that the upgrade is getting rolled up into renewed IBM software contracts. This is turning out to be a surprise to our customers as they are now forced to move to WebSphere 7.0 a bit sooner than they had expected. The good news is that Phurnace can certainly help with their migration from previous versions of WebSphere to the latest. This Phurnace-enabled migration will eliminate the need to re-craft a customer’s deployment scripts. Migration to new WebSphere versions is an “impending event” that motivates many of our customers to act.

Also, we did get some questions about properties-file based configuration in WebSphere 7.0. I directed most of those folks to our previous blog entry about it. Though it does gladden our collective hearts that IBM has addressed customers concerns about WebSphere configuration, the customers we spoke to are still not happy. The major complaint we got was about the explicit nature of the properties files and the inherent messiness of properties files. Simplicity does not seem to have been a design objective on this new feature.

Specifically, if you want to extract configuration data, you have to explicitly tell wsadmin the Object that you wish to extract. Moreover, if you extract a large amount of data, the structure of the properties file is sequential, so edit at your own peril.

Finally, we got rave reviews about our IBM WebSphere Portal 6.1 support. Give us a call and we’d love to show it to you if you missed it at the conference.

Remember, kids: friends don’t let friends write scripts.

In Untagged 
Comment (0) Read More...

Among all the questions about putting your business apps into the cloud, the one that is most perplexing is what do you do with your existing database? After all, you have an existing infrastructure. You have processes. You have a DBA resources. Why uproot all that effort for some silly Java application and this Cloud nonsense?

Well, now you don't have to. Google officially announced today that they are going to allow Java to run on their App Engine. But, just as interesting, you will be able to connect to your back office database behind your firewall using Secure Data Connector (SDC). Besides activating SDC from your control panel, you'll have to install the SDC client on your network and allow it network access to your DB. Think of it as GoToMyDB.

Of course, there will always be a DBA that will squawk about this because it's just different from the way they do things today. But, you simply can't argue with Google's commitment to eliminating objections to using their App Engine instead of hosting your own applications.

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

A few weeks ago, whurley turned me onto Puppet. Puppet is a systems management framework that allows its users to concentrate on WHAT they want to accomplish and not HOW. Moreover, Puppet configurations can be applied over-and-over again without any concern.

Obviously, I'm a huge fan of this approach as this is what Phurnace Deliver does for your Application Servers. By allowing these configurations to be applied to your servers, regardless of state, you are no longer in the business of keeping track of the state of your servers. When in doubt, just run Puppet or Phurnace Deliver with the latest configuration and your server will be "caught up". Moreover, if you have a base image that is kept up-to-date, you can then redeploy your applications and configurations using a combination of Puppet and Phurnace Deliver. This frees you from having to patch your running instance. Simply fire up another copy of your base image, install the software and patches and deploy your Java application with Phurnace Deliver. Then, you can simply use the load balancer to move new sessions to the updated server.

With this approach you get so many benefits:

• avoid downtime during updates
• automatic disaster recovery without extra effort and
• no more image sprawl and associated management hassles.

You can read more about this over at Luke Kanies' blog. He does a great job of describing this approach for handling virtual images or servers in the cloud.

In Puppet
Comment (2) Read More...