The Large-Grid Issue, Page 2
Decoding the Grid
Throughout this analysis, keep in mind what I said in the previous section, in that the first application group is more willing to deal with inefficiencies of the vendor, but this "tolerance" does not persist with the second or third application teams. As such, and indicated in Table 1, many of the Grid attributes are trivial to the first app group. They are primarily focused on early-stage execution success. To the contrary, the second and third app teams are very demanding of the Grid vendor. These users are evaluating the overall attributes of the product, and thus are more critical. Features such as resource sharing, administration, network utilization, and the like are very important to them for appropriate reasons.
Imagine a scenario where you have a couple of applications that are interested in Grid technology and looking into what it would take for them to undertake the effort. There is no central point of support, and asking the vendor of support will cost some hefty consulting dollars. The new users don't necessarily feel that they were involved in the original decision making process, and they have to somewhat live with the decision that was already made. They are looking for the path of least resistance to deployment and look for immediate ROI. This makes all the usability features very important to these folks.
Now, there are two paths that can take place here: a number of smaller pockets of Grid installations, or one Grid that is slowly growing. If one or more of the "very important" items are not met, these new users opt to 1) use a different vendor, or 2) do not join the first Grid install and create an installation of their own. Either case is not good for the organization, but the second in my opinion is worse. If multiple pockets of Grid installations take place, it gives the organization a false sense of security for moving towards the COE model that I spoke about. Ideally, you want more applications to join the Grid, and after the critical mass has been reached, move to the COE model. If multiple installations take place, a management decision will "force" these pockets to merge and that's never easy. The adoption will slow down, and you feel resistance from every newcomer that you want to join the COE.
Much of the time, an organization does not forsee the challenges ahead in Grid expansion. It is not that they are ignorant, but that they simply have not encountered the issues previously. This is why the Pre-COE era is very critical and challenging. Therefore, as you approach this scenario, make certain to select a Grid vendor that shares your vision of the future, while demonstrating flexibility for change. After all, a decade ago, having a 1000-node Grid was rare. Today, it is common and even expected for most organizations. Do everything you can to prepare organizationally, technologically, and strategically to ensure that the evolution process is as painless as possible.
About the Author
Mr. Sedighi is currently the CTO and founder of SoftModule. SoftModule is a startup company with engineering and development offices in Boston and Tel-Aviv and a Sales and Management office in New York. He is also the Chief Architect for SoftModule's Grid Appliance product that has risen from a current need in the market to manage excessive demands of computing power at a lower cost.
Before SoftModule, Mr. Sedighi held a Senior Consulting Engineer position at DataSynapse, where he designed and implemented Grid and Distributed Computing fabrics for the Fortune 500. Before DataSynapse, Mr. Sedighi spent a number of years at TIBCO Software, where he implemented high-speed messaging solutions for organizations such as the New York Stock Exchange, UBS, Credit Suisse, US Department of Energy, US Department of Defense, and many others. Mr. Sedighi received his BS in Electrical Engineering and MS in Computer Science, both from Rensselaer Polytechnic Institute.