Tuesday, May 15, 2007

Emerging from the Google OSS Grid Haze

Looks like GridGain is emerging from their semi-silence:

GridGain Systems announced today availability of the first public release of GridGain project, an enterprise open source grid computing platform for Java. This release culminates 18 months of development and presents a software product with unique set of grid computing features, open source LGPL licensing and clear focus on high-performance Java-based development.

"Almost two years ago we set out to develop a technology that would change the enterprise Java grid computing landscape in much the same way Spring and JBoss have changed the J2EE market through simplification and focusing on a developer. Today we are releasing our project that is based on proven professional open source model, business and community friendly LGPL licensing, and the host of unique grid computing features", said Nikita Ivanov, founder of GridGain project.

Tangosol (now owned by Oracle) and others gave us the Java "Data Grid". Perhaps GridGain can give us the "Compute Grid" part. This is what we struggled with in looking at data grid/caching technologies - you still needed the compute part somewhere. Tools like Tangosol excelled in bridging the gap between J2EE performance problems due to data latency issues. But that is only 1/2 of the problem in terms of application grids. You still need to be able to easily add compute nodes to be able to scale.

Without a compute grid, you need to rely on things like JMS or other bits of J2EE. This is of course fine in an evolution approach - just doesn't seem to fit right in a new large system (at least for me).

Could this be an application grid without the Jini/JavaSpaces baggage? Don't know - certainly borrows some Jini features (code mobility). They know how to document things and have a roadmap and bits so that is a start ;) Message to Apache River: see I'm not kidding, the clock is going tick, tick, tick - the future is coming.

This should be interesting to watch.

19 comments:

Cameron said...

FWIW - Nikita at GridGain has also told me that he plans to integrate with Coherence .. :-)

Peace.

Nick Morozov said...

Tangosol integration is already working in the lab and we'll be releasing it in ~4 weeks (documentation pending).

We'll be releasing intergration with Mule and WLS/WS as well.

fuzzy said...

Hey cool Cameron. Now just convince Oracle to open up Tangosol and perhaps we'll have a perfect match ;)

Cameron said...

Sure, as soon as my insurance company "opens" up my insurance ;-)

Come on, we all have to make a living somehow .. and "opening" Coherence up just means giving it away, and that doesn't make any more economic sense than giving away insurance.

Peace.

fuzzy said...

That's cool, I'd pay you for it though - I like support. I don't want it for free. I do enjoy freedom, however.

Unless you do, I'll just make due without it.

You seem to be doing fine w/o me though, but as successful as Tangosol is your time is surely limited.

At the very least, make it developer source - it isn't FOSS, but it is tolerable.

Cameron said...

I'd push to make it "developer source" if governments would provide any protections for IP. Unfortunately, we get copied already and we don't ship the source, so imagine what would happen if we did .. the world is not all nice and "fuzzy" ;-)

Peace.

fuzzy said...

I have no doubt that Oracle/Tangosol is the best distributed cache / data grid.

None.

I have no doubt that Oracle/Tangosol has the best engineers.

People can copy you all they want - once you are FOSS, so what. I'd still invest in you because I'd trust you and your engineers and you are the keeper of the bits.

Obviously this model is working for many Commercial OSS companies?

Look at it this way: You can either have (at best) another year of glory while your commercial open source competition catches up with you or you can cement Tangosol as thee Java distributed cache / data grid by acting now.

Guess what - people are going to want support for their data grid - that stuff is important. The people that want/need a data grid want support. Some of us (me) also are *petrified* of proprietary middle ware that is not based on any standard (close only counts in horse-shoes) that has severely limited switching options. Once it is open that all goes away. I don't know your business, but I would bet that net net Oracle/Tangosol would grow if it took this approach.

Thanks for the dialog Cameron.

fuzzy said...

One more thought:

You have lots of customers yes? Ok then, you already sold them licenses and you are just getting support out of them now right? Nice residual income yes? How many more new deals will you close in the next year (humor me and agree that this is all you have before some commercial OSS company catches up - say JBoss Cache).

So how about you go OSS right now. Now you are *it* - yours to lose for years. Just charge enough for support in the new model. You'll bring along all your current customers - make a little more off them perhaps and bring in the masses.

As long as it is *reasonable* it isn't about the money - really.

You could be Oracle's OSS experiment: Unbreakable Data Grid!

;)

Cameron said...

Honestly, there are simple economics at play here. You ask "Obviously this model is working for many Commercial OSS companies?"

The answer is unfortunately: "No."

Honestly, we could not sustain this business on support revenue. And if that were all the revenues that there were, Oracle wouldn't bother to buy us for it. And if companies like Oracle weren't buying companies, chances are that people wouldn't be starting companies.

I have a lot of inside knowledge of a lot of the startups and the acquisitions in the industry. The truth is a lot less rosy than it may seem.

Basing a business on the hope that people will pay for something that they really don't have to pay for is quite a gamble. I don't need my [useless] degree in economics to understand that ;-)

Peace.

fuzzy said...

"Honestly, there are simple economics at play here."?

Surely you have more experience than I do - especially based on your recent experience, but come on . . . you are going to have to be more specific to convince me. Throwing down "insider knowledge" is not really convincing.

Maybe you didn't hear me - I sometimes pay more for commercial OSS than commercial proprietary. I do this because it is better software and I like freedom: I like to see the bits, I like to know stuff. I like the freedom to fork. It keeps things real. It keeps everyone on the same page. Nobody wants to fork, but man it is good insurance.

Anyway, you clearly aren't going to open source Tangosol so enjoy your time on top. You certainly earned it. And I do wish you luck staying there. Cheers.

Erik Onnen said...

I think CP forgets how badly some OSS-based companies are crushing his fellow Oracle products. Comparatively speaking, the commercial OAS is an utter failure compared to JBoss. It's an inferior piece of engineering and Oracle does not innovate with it at anywhere near the pace that RedHat and JBoss do. Comparable installation statistics are laughable. Or perhaps he hasn't seen how badly Oracle's flagship database is losing low to mid-range database installations to MySQL and EnterpriseDB. Strangely, all of those companies seem to be doing fine with their OSS products from service, support and training, perhaps Coherence is simply to niche to reach the size of these successes.

Cameron said...

Fuzzy: I sometimes pay more for commercial OSS than commercial proprietary.

OK, how many contracts with a value over $100k/year does your company have with "commercial open source" companies? How many over $1mm/year?

Peace.

Cameron said...

Erik: I think CP forgets how badly some OSS-based companies are crushing his fellow Oracle products. Comparatively speaking, the commercial OAS is an utter failure compared to JBoss.

I was referring to the economics, not the technological aspects. Having personally helped to put dozens of JBoss-based apps into production all around the world, I know the pros and cons of the technology. Having been frustrated with previous incarnations of OAS, I know the pros and cons of that technology as well (or at least earlier versions of it).

But looking at the economics ..

In dollar terms, Red Hat expects on all JBoss related revenue to gross about $30mm this fiscal year.

So even a little niche product like Coherence will gross more this year, and will obviously net more ..

Peace.

Nick Morozov said...

I think asking to OS Coherence is rather unrealistic. Closed source to open source transition is viewed almost universally as a failed business' last throw (e.g. Terracotta). Tangosol got started in a lucky time when people did pay for a license on good technology much more willingly than today. And they have managed to stay on top despite heavy competition (at times :-).

I absolutely certain that if Cameron were to start Tangosol today things will be different. Professional OSS business model is maturing almost daily nowadays. Evaluations go up, deals get made, etc. Look at Jboss, Spring, Geronimo, etc. etc. Dynamics of these deals are different, the expectations are different (longer terms) but fundamentally the middleware market has really shifted towards professional OSS.

fuzzy said...

You are probably right Nikita.

I agree with your description (last gasp), but as commercial oss becomes more popular this won't always be the case. There certainly will be good products that change business models. That is of course if the math is good or bad enough, switching costs are low, etc.

I have zero hope of Oracle even allowing it - let alone Cameron recommending changing to commercial oss.

The conversation started in jest really. Cameron is a bit of a blog troll. He is quite fond of dropping in on blogs, TSS threads to market his product etc. (smart man). So I hazed him a bit for fun. I really do like his product - I just am petrified of proprietary non-standard middleware.

Cameron often doesn't really take the conversation anywhere when he trolls. He usually just swats at the competition and then when someone asks a real question, he just abandons the thread. This time Cameron actually said some interesting stuff :) Thanks Cameron.

fuzzy said...

Cameron: "In dollar terms, Red Hat expects on all JBoss related revenue to gross about $30mm this fiscal year."

Why did RedHat spend over 10x on it then ($350MM)?

Cameron said...

Fuzzy: Why did RedHat spend over 10x on it then ($350MM)?

It was a lot more than 10x, since the multiplier is based on the previous 12 months before the acquisition (i.e. the $30mm is what they are getting two years later!)

To be honest, it's a good question why they felt they could pay so much for so little, especially considering that after purchasing it they lost the entire JBoss management team (a very talented group of people that you've never heard of) and a number of the engineers.

BTW - The original RedHat estimate for FY07 was $80 million in JBoss revenues, so they have obviously cut it down by quite a bit! See:

http://woodrow.typepad.com/the_ponderings_of_woodrow/2006/04/jboss_gets_acqu.html

It was a very poor investment decision, and the RedHat stock has never recovered from it. With the rest of the market going "up, up and away", RedHat somehow managed to find something to buy that could drag them down and keep them there.

$.02 .. FWIW .. etc.

Peace.

fuzzy said...

Cameron: Wow I like the new blogsphere Cameron. Thanks for the thoughtful discussion :)

Time will tell I suppose. Are you referring to Marc Fleury or some of the people behind him? He certainly is brilliant and is a loss. I'm glad to see him blogging. He has an important voice.

As for RedHat, a lot of people have bet against them - I don't think we have heard the end of JBoss by any stretch. RedHat may have paid too much (lots of VCs needed paying?), but JBoss has quite a brand and is definitely trending up.

Like I said, I am not a free loader. Cost is not the primary reason why I like OSS. If there is cost savings for the bits, great. I just like the flexibility, community, innovation, etc. You can't get that with traditional commercial proprietary.

I hope that as the commercial oss model evolves people will come up with more ways to monetize it etc. Short of a nuclear patent war, I don't see how commercial oss doesn't continue to grow.

Cameron said...

Fuzzy: Are you referring to Marc Fleury or some of the people behind him?

There was a really good management team behind Marc.

Fuzzy: As for RedHat, a lot of people have bet against them - I don't think we have heard the end of JBoss by any stretch. RedHat may have paid too much (lots of VCs needed paying?), but JBoss has quite a brand and is definitely trending up.

I think RedHat is doing fine, and the JBoss group at RedHat is doing fine.

I was only looking at the economics of it. They call it "ROI" for a reason ;-)

Fuzzy: I just like the flexibility, community, innovation, etc. You can't get that with traditional commercial proprietary.

Trust me, if there were ways to safely provide source for our product, we would do it. There are so many benefits to it. We do have a great community around our software (e.g. we had 70 people at our London user group a few weeks ago, and check out our online forums), but I know I personally like to have source when I'm developing (or worse yet, debugging!) against a product.

Peace.