Ruby, COBOL jump on Amazon cloud
Or were they pushed?
The Vertebra project is about creating tools that help manage the deployment of applications on and across clouds. It is arguably a first step, of course. But given that the code behind Vertebra has been released under the GNU Lesser GPL open source license, there is now an opportunity for it to be embraced by the open source community and extended.
COBOL compiler and legacy application modernization tool maker Micro Focus also wants to get into the cloud racket, saying this week that it has tweaked its Micro Focus Server, the runtime environment at the heart of its COBOL tools for mainframe, Unix, Linux, and Windows platforms, so it can be deployed on the Amazon EC2 and S3 utilities.
"We're surrounded by 17 Google buildings in our Mountain View offices, and we figured that if we can't sneak into Google's free cafeteria, we'd better get out there on the clouds," quips Mark Haynie, chief technology officer at the company's application modernization unit.
Like Engine Yard, Micro Focus knows the smart move is to be cloud-neutral, and while the COBOL tool maker is delivering support for COBOL applications on EC2 and S3 today, it has every intention of supporting Microsoft's Azure cloud (and has tested on the prototype for it). The interesting bit is that Micro Focus is guaranteeing its customers that COBOL applications have been sufficiently abstracted by its software so they do not have to tweak them at all.
For instance, COBOL applications that have VSAM files for keyed access reads still work. The VSAM files are, in fact, stored in an SQL database as a BLOB out there on the cloud, and because of the tweaks in the Micro Focus Server, the COBOL applications are none the wiser. Even Microsoft is not promising such compatibility with its own Azure cloud, according to Haynie. If you have C# applications, you have to tweak them so they speak to Azure services instead of Active Directory services on a local machine.
Micro Focus expects to deliver its Amazon cloud variant in May at its annual user group meeting. For now, only early adopters are getting their hands on it, and in May, the company will show off their experience and do a formal launch, complete with pricing. While not being specific about pricing, Haynie says it will be a subscription-based model, perhaps with a fee per month per registered user or per hour of compute time used. ®
Sponsored: Beyond the Data Frontier