This blog post de­scribes yet again how you can use best-of-breed com­po­nents to do the job.

In par­tic­u­lar, it de­scribes in­te­gra­tion of Atomikos as well as all of the oth­er cool li­braries that to­geth­er out­per­form the ap­pli­ca­tion serv­er in sim­plic­i­ty, cost and per­for­mance... Good job!

The re­cent an­nounce­ment of Ap­ple's cloud il­lus­trates yet again that a new era is about to start: that of the cloud - and with that, the mas­sive adop­tion of ser­vice ori­ent­ed ar­chi­tec­tures (SOA).

This will have an im­pact so huge that it is hard to un­der­es­ti­mate. Some of the ob­vi­ous changes that will hap­pen are most like­ly the fol­low­ing:

• In­fras­truc­ture (op­er­a­tions) will be­come ag­ile - at the ex­pense of IT jobs in op­er­a­tions. Ap­pli­ca­tions will be de­ployed with un­prece­dent­ed ease and speed. This is be­cause the cloud in­fra­struc­ture will trans­par­ent­ly al­lo­cate the re­sources that now are al­lo­cat­ed by hu­man ad­min­is­tra­tors.
• SOA will be­come main­stream. This is be­cause even the most com­mon in­fra­struc­ture ser­vices will be vir­tu­al­ized, run­ning in the same or a dif­fer­ent VM from the ap­pli­ca­tion, and scal­ing up and down as need­ed.

The in­ter­est­ing part - at least to us - is that trans­ac­tions will be­come more dis­trib­uted than ever. We are ready for this change - are you?

Data con­sis­ten­cy in the cloud is a to­tal­ly dif­fer­ent ball game - as per­fect­ly il­lus­trat­ed by this newslet­ter is­sue from ZapThink. Essen­tial­ly, you have to cope with some form of par­ti­tion tol­er­ance and stale data val­ues in caches. This is usu­al­ly sum­ma­rized as the CAP is­sue (for more de­tails, see the ZapThink post).

Luck­i­ly, the so­lu­tion to the prob­lem can be as sim­ple as out­lined in our so­lu­tion to the CAP prob­lem.

Life is good

The cloud is an in­ter­est­ing phe­nom­e­non, to say the least. Just last week, I wit­nessed a demo by Adri­an Coly­er at the What's Next (Paris) event that left me very im­pressed be­cause I could eas­i­ly un­der­stand the prob­lems that it solves:

• Slow in­fra­struc­ture set­up is re­placed by in­stant avail­abil­i­ty of re­sources (data­bas­es, hard­ware)
• Scal­a­bil­i­ty is triv­ial­ly easy: on com­mand is all it takes
• Ser­vices are pro­mot­ed as the means of reuse - ef­fec­tive­ly en­abling SOA strate­gies

I know from ex­pe­ri­ence that these are all pain points in to­day's IT in­fra­struc­ture for just about every big or­ga­ni­za­tion. So for main­stream op­er­a­tions, I think the cloud is go­ing to be a killer.

While this is all fine, I still won­der how this will work out for one of our core mar­kets: fi­nance. Sure enough, ini­tial of­fer­ings are show­ing up, but so far noth­ing re­al­ly in­ter­est­ing for our re­tail bank­ing au­di­ence. Add to this the fact that most banks are typ­i­cal­ly very con­ser­v­a­tive (they have to be) and the cloud for the bank­ing world seems a bit more of a fu­ture sit­u­a­tion than for the rest of the in­dus­try.

But cer­tain­ly that time will come too, and for those days Atomikos has the ex­act kind of tech­nol­o­gy for cloud-en­abling trans­ac­tion guar­an­tees that banks de­pend on so much. That is be­cause we of­fer the third gen­er­a­tion of TP mon­i­tor­ing tech­nol­o­gy: light-weight, em­bed­d­a­ble, per­fect­ly scal­able and SOA ca­pa­ble too.

The fu­ture is bright;-)

Check out this in­ter­est­ing pre­sen­ta­tion on In­foQ by Ce­sare Pau­tas­so.

We at Atomikos are work­ing with Ce­sare to re­search REST and Trans­ac­tions and how to in­te­grate this with BPM for REST as well; check out this slideshare pre­sen­ta­tion (also by Ce­sare).