Cloud Foundry: Design and Architecture
Derek Collison discusses the goals, the design premises and patterns employed in creating the architecture of Cloud Foundry, VMware’s open source PaaS, unveiling internal architectural details.
The content has been bookmarked!
There was an error bookmarking this content! Please retry.
Posted by Scott Delap on Sep 06, 2006
In the first of a series of articles on continuous integration Paul Duvall compares three popular continuous integration servers, Continuum, CruiseControl, and Luntbuild. He mentions that there are a large number of commercial and open source CI servers currently available. Each has its pros and cons which makes picking one a difficult task. He recommends considering the following criteria:
The article then walks through installing and running each CI server being compared. Paul provides different recommendations depending on your goals. He rates Continuum as the easiest to use. CruiseControl is the most extensible with a thriving user community. Luntbuild offers extended support options.
Introducing SQLFire: a memory-optimized, high performance SQL database
Agile Maturity Model Applied to Building and Releasing Software
Continuous Delivery: Anatomy of a Deployment Pipeline
Adopting Git for the Enterprise: Risks and Considerations
Branching & Merging Efficiently: A Guide to Using Process-Based Promotional Patterns
VMware vFabric SQLFire - Test drive the data management system with memory speed, horizontal scalability and a familiar SQL interface
I'd agree that Continuum is easy to use, but it seems to force strange defaults. CruiseControl is definitely more difficult to learn, but once you've tackled the learning curve - it's great (IMHO of course).
I'd agree that Continuum is easy to use, but it seems to force strange defaults. CruiseControl is definitely more difficult to learn, but once you've tackled the learning curve - it's great (IMHO of course).
I agree with you, we use Maven 2 as our build system and continuum has a lot of very annoying bugs at the moment.
For instance, if you try to remove a project and click somewhere else before the request is completed (double submit), the project is still going to be in the list but you won't be able to delete or build it anymore. The only solution I have found so far is to use a database client and dump the project data.
Also when you add a multi-modules project, every sub-modules are added by default. If you prefer to build the project as a whol,e you have to remove all of them by hand. It takes some times and if you are not careful, you are probably going to get the first bug I described.
On the other hand, I have to say that the project is still in his infancy and it does look very promising, it is very intuitive to use and has some very nice features. Hopefully, the 1.1 version should be more stable.
We've been using it for an 'enterprise' project with multiple sub-projects on the same lifecycle, and Cruise isn't well-suited for that. For instance, while it can have a 'parent' project trigger a child, it can't easily have a child not build if the parent has unbuilt changes.
We've been able to create a plugin to do that for our own purposes, but at the moment, each build check goes against CVS, which means that the 'core' projects get cvs-checked extremely frequently, which could be a bottleneck on the build.
Luntbuild is pretty easy to setup and has a lot of nice features. Build dependencies going either forward or backward like you're talking about is one of them. It's also nice that it's a complete build management app, not just continuous integration. You can have your QA department do their builds from Luntbuild and download the resulting binary. They can test that build and, if it passes, they can promote it to be the "gold" release and use Luntbuild as the build repository.
The UI takes a little getting used to, but once you're familiar with it it's pretty quick and easy to get around and get things done. I haven't tried the Maven support... I've never been a big fan of voodoo, so I stick with Ant.
Yes, Luntbuild is the one I'd been leaning towards experimenting with; had some troubles getting it up and running on the build server either under Resin or with the bundled Jetty, and I haven't tried again yet.
I've used CC on a previous project, it was very good. We were using ant for builds and the combination of the two worked very well. The learning curve w/ CC isnt that bad, but there is definitely a curve. An ant task had to be created for integration w/ Accurev SCM since there wasnt one out of the box.
On the latest project we opted for Maven 2 (which I really like). After doing some research we chose Luntbuild. Luntbuild's Maven 2 support is great. We are using it on a multi-module project with 10+ modules. Setup was a snap, luntbuild was up and running our Maven build from ClearCase in under an hour. I highly recommend luntbuild for its ease of use/administration. Although I have not used the Luntbuild/Accurev integration its good to know that no custom code would need to be written for Accurev use.
Derek Collison discusses the goals, the design premises and patterns employed in creating the architecture of Cloud Foundry, VMware’s open source PaaS, unveiling internal architectural details.
Andrew Watson talks about the work of the OMG, where CORBA is alive and well (hint: in your car), UML and UML Profiles vs. custom Modeling languages, DDS and other middleware, and much more.
Sohil Shah discusses creating iPhone and Android enterprise mobile applications based on cloud services using the open source platform OpenMobster.
Paul Sanford presents the transformations supported by data throughout its life cycle, and how that can be better done with Splunk, an engine for monitoring and analyzing machine-generated data.
A common “best practice” for unit tests is to only write a one assertion in each test. I intend to question this advice by showing that multiple assertions per test are both necessary and beneficial.
John Rauser presents the architectural and technological evolution of Amazon retail websites starting with 1994 and ending with adopting Amazon Web Services.
Michael Stal discusses system architecture quality, how to avoid architectural erosion, how to deal with refactoring, and design principles for architecture evolution.
Every developer has had to integrate with another system, API or component. Tis article provides strategies to handle the change and for he separating system boundaries.
6 comments
Watch Thread Reply