Trilateral Agreement - The Impossible Happened!
As I mentioned before, I'm working at a sizable Federal Agency in the Washington DC area. The contract is big and I'm a small fish on this farm.
During the last month or two, several of the contractors have brought in teams of webMethods people. Some are obviously more advanced than others. This is the dilemma.
There is nothing better than direct experience with webMethods.
Very experienced webMethods consultants usually have an uphill battle with the other teams trying to gain buy-in to various architecture components.
Having several years of java experience helps, but doesn't prepare one to advise on large integration projects. Ditto for ASP, PHP, .net and other technologies. I have senior .net architects dictating to me architecture based on something other than my experience can relate to.
SO, in my uphill battle to reduce the number of integration points, packages, etc., I have this serious uphill battle.
Then, yesterday, the impossible happened. Either I wore the senior architect for the project down, or he had a mental collapse. He agreed to everything that I wanted to do across the board without comment.
With this agreement, I turned to the other two contractor groups who immediately got on the bus. Trilateral Agreement!
With glee, I returned to my cube, revamped my code to support (yeah, you guessed it) enterprise-wide transport documents. So, rather than having a request and response document for each message, I will have ONLY ONE request and ONE response document for the entire sysetm.
And, instead of having an interface for each document, I will have a single interface into the system. This interface will determine payload and redirect the request to internal flow services.
The long-term benefits include faster integration of future messages, easier debugging, and less-complex security models. The current project does not include the use of TN or Broker, but now I can say without any guilt that it would take MINIMAL efforts to institute broker messaging and a moderate amount of effort to replace the single integration point with Trading Networks.
One Notch on my belt.
During the last month or two, several of the contractors have brought in teams of webMethods people. Some are obviously more advanced than others. This is the dilemma.
There is nothing better than direct experience with webMethods.
Very experienced webMethods consultants usually have an uphill battle with the other teams trying to gain buy-in to various architecture components.
Having several years of java experience helps, but doesn't prepare one to advise on large integration projects. Ditto for ASP, PHP, .net and other technologies. I have senior .net architects dictating to me architecture based on something other than my experience can relate to.
SO, in my uphill battle to reduce the number of integration points, packages, etc., I have this serious uphill battle.
Then, yesterday, the impossible happened. Either I wore the senior architect for the project down, or he had a mental collapse. He agreed to everything that I wanted to do across the board without comment.
With this agreement, I turned to the other two contractor groups who immediately got on the bus. Trilateral Agreement!
With glee, I returned to my cube, revamped my code to support (yeah, you guessed it) enterprise-wide transport documents. So, rather than having a request and response document for each message, I will have ONLY ONE request and ONE response document for the entire sysetm.
And, instead of having an interface for each document, I will have a single interface into the system. This interface will determine payload and redirect the request to internal flow services.
The long-term benefits include faster integration of future messages, easier debugging, and less-complex security models. The current project does not include the use of TN or Broker, but now I can say without any guilt that it would take MINIMAL efforts to institute broker messaging and a moderate amount of effort to replace the single integration point with Trading Networks.
One Notch on my belt.
posted by Ray Moser at 1:12:00 PM
1 Comments:
Good job. It looks like your experience and authority in integration is finally being recognised. There is always an up hill battle until the powers that be decide that your ideas resonate. Some times it never happens because of ego battles and that's when it can be time to avoid a disaster and head for the door.
Post a Comment
<< Home