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 Charles Humble on Oct 19, 2010
Replay Solutions announced today that Larry Lunetta, formerly of ArcSight, has joined the company as president, CEO and member of the board. During his seven-year tenure at ArcSight, Lunetta helped launch the company, now publicly trading with more than $100 million in annual sales. Lunetta has also worked at the executive level at several start-up and public Silicon Valley companies, including ePAC Technologies and Branders.com.
Lunetta replaces Replay's co-founder Jonathan Lindo, who will continue to lead development and partnerships with Replay Solutions as executive vice president of products and technology. Lindo told us:
Larry has deep experience in taking new technologies into established markets. His background in developing and growing emerging companies will allow Replay Solutions to take advantage of not only the need to produce higher-quality code, faster in response to changing business requirements, but at the same time add to the security and compliance of key applications by spotlighting hard-to-find flaws and bugs that are tempting targets for malware.
Replay Solutions' core product - ReplayDIRECTOR for JavaEE - is a software debugging tool that can be used to capture events that occur in a system running in production, allowing them to be subsequently replayed in a virtualised environment to identify the root cause of the problem. Running the product does incur some overhead, but the firm claims it is small – in the region of 1-5%. Lindo explained that the product
...works by capturing, recording and replaying non-deterministic events and inputs. We are using byte code instrumentation at class load time to analyze when non-deterministic data can enter the application, and adding a small amount of byte code to capture and record the data during run-time.
We are using a small handful of the JVM TI methods in order to instrument the application at load-time. We actually don't use redefineClass(), but perform instrumentation with other features provided by JVM TI and the run-time.
One of the primary reasons we are able to keep overhead low is that we only are concerned with a very small percentage of the data flowing through an application. Because most code execution is deterministic, our system simply ignores it, making the overhead low.
We discussed why the firm built a custom virtualisation solution, rather than building on top something like VMware's Hypervisor. Lindo told us
We built our application virtualization from the ground up to achieve the level of performance and flexibility required to deliver a seamless and transparent solution. Most virtualization technologies operate at the system level, like VMware's Hypervisor. We virtualize at the application layer allowing ReplayDIRECTOR to isolate the application execution from the OS and hardware on Replay. This is what allows us to replay and reproduce any application issue on a different machine, regardless as to whether that machine is virtual or physical.
ReplayDIRECTOR works very well with system-level virtualization products including VMware, Microsoft Virtualization or Citrix Xen.
ReplayDIRECTOR faces competition from a number of larger vendors such as VMware, Microsoft and BMC. VMware has a record/replay feature in their Workstation product, Microsoft offers IntelliTrace with Visual Studio 2010, and BMC has a product called AppSight. All the products look to address the same problem, but according to Replay their product is unique in that it is able to isolate and record just the application. This in turn allows a developer to re-run the sequence that triggered the bug, providing a means of retrospectively re-creating a production problem in a debugger.
ReplayDIRECTOR requires an agent to be installed on the application server, and is currently certified for Tomcat, JBoss and WebLogic on Windows, Linux and Solaris. It is offered in both a hosted environment which is free for a single JVM rising to $3,800/month for 20 JVMs, or as a slightly cheaper stand-alone download at around $3,200/month for 20 JVMs. It can be downloaded for free from the company’s website.
Introducing SQLFire: a memory-optimized, high performance SQL database
Troubleshoot Java/.NET performance while getting full visibility in production
RDBMS to NoSQL: Managing the Transition
Combining Inspections, Static Analysis, Testing to Achieve >95% Defect Removal Efficiency
VMware vFabric SQLFire - Test drive the data management system with memory speed, horizontal scalability and a familiar SQL interface
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.
No comments
Watch Thread Reply