bavads

jmxterm

2011-03-16T18:35:00.001-07:00

one of the best utilities i have used... took me just a couple of mins to get started....

very helpful if you have to write your own monitoring of your applications using mbeans .. easily coupled with rrd and drraw it is a breeze... although you have tools like hyperic ... et al .. there is no greater fun than writing your monitoring solution...

more info once i have a readme on this... a simple how to..

using the following command - you can get the queue for the JMS topic

echo get -s -b org.apache.activemq:BrokerName=localhost,Destination=TASK.EVENT,Type=Topic QueueSize | java -jar jmxterm-1.0-alpha-4-uber.jar -l localhost:7010 -v silent -n

start your app with arguments like -Dcom.sun.management.jmxremote.port=7009 -Dcom.sun.management.jmxremote.ssl=false -Dcom.sun.management.jmxremote.authenticate=false

JAVA VM monitoring via snmp via jmx

2010-12-03T11:58:00.000-08:00

Never let go of this link
http://docs.sun.com/app/docs/doc/820-5704/bhamf?l=en&a=view

also complete the information here for the entire set-up of monitoring using snmp jmx rrd drraw

Mac, eclipse, cocoa, netbeans and some more

2010-10-07T20:57:00.000-07:00

Eclipse using cocoa on mac os x gave me the biggest headache...

could not cut, copy, paste in eclipse. nothing got copied to the clipboard and no help on the net to resolve this issue. after spending a lot of time, gave up and installed net beans. Am impressed with Netbeans, it is light, sleek and impressive. However, importing eclipse projects into netbeans isnt clean enough and hence gave up on getting everything right. too much initial investment for the move from eclipse to netbeans.

After many tries, got the EasyEclipse distro which is good and allows COPY & PASTE without any issues...

so far so good ... till the next blocking issue...

====

figured out the entire mess ...

I was starting eclipse from shell and this caused all the problems. If you start eclipse as such, the paste command is not directed to the UI, but is pasted to the shell...

I started eclipse from the dock and everything worked ok... i had to own all the files in the workspace / projects and that was it... now copy / paste is no longer a problem...

Inspiring

2009-07-29T11:08:00.001-07:00

Looking around, we find so many things that are inspiring .... Read this update and looked at the things that engineering teams brag about ...

http://blog.jinspired.com/

This kind of information just inspires you to reach new heights instead of being bogged down by the difficult times ....

Marc Andreessen - VC and Platform

2009-07-07T10:44:00.000-07:00

Heard about this yesterday and was pleasantly surprised by the news bit on NPR..

http://www.businesswire.com/portal/site/home/permalink/?ndmViewId=news_view&newsId=20090706005222&newsLang=en

Platforms explained...

http://blog.pmarca.com/2007/09/the-three-kinds.html

I like his blog ... the clarity of thoughts and the optimism of approach ..

Platform ...

2009-07-01T12:07:00.000-07:00

My thoughts that are exactly echoed here ...

http://longjump.com/platform/platform-features.htm

What I feel is that platform and services around it are all combined...(must be for marketing)... but I believe platform should provide just the basic features - rather be only the runtime and should address only few concerns ...

Interesting to read this though ... for me .. just one more validation of my thought process..

Is JSON the way

2009-06-30T12:56:00.000-07:00

Object Notation, dynamic proxy, plugin architecture - Have been thinking about these two a lot nowadays... in the context of I ... my nickname for the platform. I = infinitely scalable...

Is it possible to build a platform that provides this capability ... ? Also can the same paradigm be used for web framework...

An anology from a case study done during my business school days ... about boeing... the thing that struck me is the way planes are built ... this is not an assembly line concept where work passes through stages ... but plane (an object, artifact) being built where people / processes collaborate to get a unit of work done.

How can this be applied to a platform that is built around similar concepts where unit of work defines the collaborators. If this concern can be addressed then what you get is an infinitely scalable platform. With a simple jdk (any runtime for that matter) and a farm of servers can the platform orchestrate this unit of work / delegate to a server who is free and provide the necessary collaborators for completing the same.

Declarative definitions for a model and a technology like JSON which is a simple object definition, can they come together, be dispatchable to a free server and a unit of work be done and can it be done fast and reliable. The notion of a virtual platform which manages a pool of workers who have nothing but a jdk (or any runtime) seems appealing.

This paradigm can also be then applied to a web framework where you dont write tons of jsps / asp / php ... but platform provides the necessary objects & templates and instructions to get a unit of work done (say render a widget or a portlet or a page)...

Idle mind .. is a dreamer's workshop...

Need to get a job first ... my 2 week job sprint (finding a job sprint) going on.... !!!

Pearson getting a stake in a couple of Indian Companies

2009-06-25T09:12:00.000-07:00

Interesting to see this news....

http://www.nytimes.com/2009/06/25/business/global/25rupee.html?_r=1&ref=technology

Starting your own gig

2009-06-23T10:19:00.000-07:00

Last two weeks have been interesting. Talking to lot of people and running my ideas around, I got good / interesting advice. Let me hash out a cookbook (cheat sheet) for anyone who is interested and jot down notes that I can come back to later.

1. VISION

Cliched but essential. Not going down to the grand description as taught in business schools. What is needed here is a definition of what you want to achieve. Right now I have an idea which is I would say is ummm..... half baked. It is in my mind constantly and I keep thinking about it day and night. However, what I need to do at this juncture is verbalize it. A good bit of advice given to me is put it down in a powerpoint. This achieves two goals - a. Helps verbalize ideas in your mind & b. this is the medium by which you can attract / recruit a team who would contribute their time.

2. BUSINESS PLAN

Again not a grand treatise here, but penning down series of progression on what need you are trying to resolve. Need, Solution, Competing solutions, Strengths of your solution, Adoption (sales) cycle, Possible monetiziation.

At this point, you have a basic set defined of what you are embarking on. Next is getting feedback. Talk to people and discuss. In general, you tend to have a bias and you need someone to challenge your thoughts.

3. TEAM

After this go about getting good people on board. Work with a team of 5-6 to contribute towards a mashup / prototype. This would be a new experience for me and I know the challenge would be managing engineers who are innovators and working on their own time with no assured returns.

My ideal team that I am looking for is

1. My Self in the role of a product manager (define vision), project manager(timelines/dependencies), designer (architect), code (developer), test (QA), SCM, IT, Janitor ... all in one.
2. Technologist - An architect who is focussed on simplicity and elegance.
3. Advisor - Thought leader necessary to expand the boundaries of current thinking.
4. Developer1 - Front end developer
5. UI designer - Images, flash, css, look and feel

4. PROTOTYPE & ADOPTION

Build a simple prototype (but i do want it jazzy to get the wow factor) with limited functionality.

Deploying your proto and getting some traction I feel is essential to get feedback on what you are doing. Start sowing seeds of this thought process right now. The points mentioned above are all under your control. This is the point of inflexion where you are ready to show your solution to your target. I believe much will depend on how this phase goes.

5. FUNDING

I am not clear on this aspect, but this would be taking the next step in your enterprise. There are variety of avenues that can be used depending on what you need. Angel / VC / equity / debt et al. But here I feel an understanding of what you need, how much you need and why you need needs to be hashed out. As a general principle what I feel is that this negotiation needs to be at a level playing field. As much as you need funds, I believe in this economy, people with money also would be looking out for ventures that they can invest in.

But I will think about this around the time I feel we are there.

I dont know how I am going to fare, but one thing came to my mind as I was talking to a friend of mine recently - "I dont want to be accused of thinking small and failing, but I want to think big and succeed."

Mock Objects

2009-06-18T09:53:00.000-07:00

TDD and Mock Objects

Been playing with this for some time now. Test Driven Development requires a change in approach on designing and developing code. My natural instincts are to understand the requirement, design contracts (interfaces), implement and then unit test them. TDD disrupts the natural cadence and adapting to this approach requires some diligence and discipline.

I have not entirely discarded this approach but read an interesting interview by Donald Knuth - http://www.informit.com/articles/article.aspx?p=1193856

So coming back to Mock Objects - here are my 2 cents. I feel that it makes sense to Mock Objects at early stage of development during the time the code is getting baked in. Some of things become difficult when you play with Mocks. For e.g. a method that does some heavy business logic where multiple objects collaborate, you end up mocking a lot of those. The test becomes that much more complex and you end up writing more code. Then there are frameworks... I looked at jmock and EasyMock. I discarded jmock soon when I realized that I cannot mock a class without an interface. I dont want to write interface for every object that I want to use. EasyMock gives a good classextension which is handy to mock classes and specific methods. However, what I found out this also needs methods to get/set the objects to mock.

For e.g.

UserService{
authenticate(...){
UserDao ud = new UserDao(); // Cannot mock UserDao which is instantiated within the method
ud.fetch(...);

}
}

So I end up writing a getter / setter method and then mock it and set it to what I want UserDao to do.

I can easily write a unit test that actually works with a sample db and provides me data driven simulation of all the logic. Isnt that what I am actually looking for? Isnt simplicity the overall concern of any code that you write and do you want to disturb this cadence by overly complicating without gaining much...

Platform

2009-06-16T15:58:00.000-07:00

looking at things to do for my first gig, i have decided to build yet another social network. Is it difficult? ... maybe... how to monetize it? ... ad based ... huh ... not so much. people have done that and is so yesterday.

however for my quest here, i need a good foundation. so looking at platform to build a network on. and that is the basic question. what is a platform and what should it do? what should the platform provide that differentiates a platform from the services and application built on top of it.

a platform is a runtime on which models interact to provide a business solution. does it then orchestrate persistence, state, sequence of events, notifications? or is it a means to declarative define all these? right now my search is to isolate what is and what is not a platform....