Open Source Weblog

Few could doubt IBM's commitment to open source (apart, maybe from JBoss's Marc Fleury) and or the pro-open source leanings of IBM software's worldwide Linux strategy manager, Adam Jollans.

It was interesting to hear his opinion that open source will struggle to make an impact on the systems management space due to the large number of existing tools that it will have to integrate with, and/or displace.

"In terms of systems management, we're not yet seeing a lot of adoption of open source management tools as yet," he told me. "It's harder in terms of systems management because you've got a lot of existing tools there. The question is, does this provide a better solution."

A fair point it seemed, and a potential explanation as to why open source systems management might take some time to make an impact. That was, until I spoke to Fred Gallagher, VP of marketing and business development, and William Hurley, chief technology officer, of Qlusters, which launched its openQRM project on SourceForge in January.

According to Qlusters, the adoption of open source systems management tools will be driven by that very need to provide integration with the various tools that are out there. "Data centers are unique, customers already have tools and need it to work with them," explained Gallagher.

"By going open source the theory was we can get users involved to develop that integration and hire third party developers to develop that integration," he added. "For systems management to work there needs to be more open source services," maintained Hurley.

Hurley is convinced that open source is not only welcome at the systems management level but absolutely critical if companies are to tie together the various vendor-developed and homegrown scripts and tools and achieve the interoperability that systems management framework vendors have promised but failed to deliver.

Hurely added that he has been convinced of this fact since he was involved in research about the future of the systems management software market while at… IBM Tivoli. The research indicated that open source was the way forward, Hurley said, but IBM "didn't do anything" with it.

While the openQRM project has got off to a good start it is still too early to say whether Jollans or Hurley will be proved correct, but I couldn't help be reminded of comments made by Oracle, BEA and IBM just under two years ago about the potential for open source in the middleware and database markets.

"Are IBM going to invest in MySQL? Are Oracle? Who's going to invest in it? There is no other way to develop the improvements required," maintained Oracle's EMEA product director, Andy Cleverly in 2004, which makes for interesting reading given Oracle's acquisition of Sleepycat and reported attempt to acquire MySQL itself.

"We're not seeing any impact here," said Surjit Chana, IBM VP of software marketing strategy and planning of the potential for impact on its DB2 and WebSphere products. "In very low-end pieces of the market we've seen some of the open source products playing. Those are very small, niche focused areas so we don't see as much momentum as we do with Linux."

The momentum had picked up enough for IBM to decide to purchase Gluecode a year later in May 2005 and launch WebSphere Community Edition.

All of which does not guarantee open source's success in systems management. But the signs are there.