To: Raymond Duray who wrote (1943 ) 2/5/2001 3:02:11 PM From: Frank A. Coluccio Read Replies (1) | Respond to of 46821 Hi Ray, Yes, I see a leader in this space. One appears about once every three years. You can usually tell who they are by the sticks with feathers on them protruding from their backs. From experiences I have had with previous such third party applications, there is a tremendous amount of customization required with each vendor's specific architecture. For larger carrier architectures, we're talking Manhattan Project. There are industry standards in varying levels of refinement that are supposed to facilitate this task, such as those which use object code and corresponding APIs, but adherence to these has been less than stellar. Especially among startups (those who have come onto the scene in the past two or three years) who have all to do simply to get their platforms up and working. Recall several such initiatives? Bellcore's OSMINE? The ITU/ANSI -sanctioned TMN (Telecommunications Management Network Framework)? Open Systems Management Protocols? Simple Network Management Protocol (SNMP) MIBs? SNMP doesn't have all of the hooks that are required of "carrier class" devices by service providers (mostly the entrenched, legacy ones), and many vendors don't adhere to Common Network Management Protocols (OSI). So, a manager of manager (MOM) system must, by definition, be able to handle both at the subsystem and element levels. Version controls for each? Fudgeddaboudit. "Third Party" point and click solutions are achieved across a broad variety of platforms only after a lot of hard work, if ever at all in a universal sense. In the context of out-of-the-box "any vendor" functionality, it remains a holy grail of sorts, and at best a euphemism, when one considers all of the arduous behind-the-scenes work that is most often involved to get it to work. Another aspect to consider is that once you get beyond the first network element you must orchestrate its behavior with others through operations support systems links (OSSes). This requires a Leonard Bernstein approach to writing score, but there is no guarantee that you will always wind up with something sounding like Maria. It's doable and it is being done every day, but it often doesn't scale well unless lots of room has been left in the design, making additional customization possible as new elements are added or removed from the carriers architecture. Nor is it necessarily portable to other platforms of the same generic, which are used in different contexts, once it is has been done for the first. Job security for programmers comes to mind here. Which isn't such a bad thing these days. Also, it's little wonder why a great many enterprises, as well as a fair number of upstart SPs, have recently turned to a single vendor approach to building there intranets and carrier platforms, respectively. FAC