My employer is a large Swiss Telco. We have many Systems used to transfer data for different tasks, e.g. Performance Management, Fault Management, Configuration Management etc.
In order explain to "Management" (pointy haired, and other) how these systems interact, I collected information about data flow/formats/protocols into a "database" ( of the comma delimited persuason) and then generated code for Graphviz (http://www.graphviz.org/) and Yed (http://www.yworks.com/en/products_yed_about.html) to visualize these Graphs.
Showing these graphs generated from my DB was rather effective at first ..
but adding new systems/data flows causes both Graphviz and Yed to re-layout the graphs, since that is how they work ...
Naturally this really confuses anyone trying to match that graph seen yesterday, to the graph shown today.
How would you display/layout Data-Flow between Enterprise Applications?
If you want to stick with graphviz and generating from your data
- Make the new graph
- Generate the Dot file with the exact locations
- Print out new graph
- Edit the dot file to remove the things that are new
- Print out the graph for previous
They can't compare the new graph with the one you gave them, but they will have two that they can compare with each other.
There seems to be no tool that does what I want.
We (=my employer, I've moved to other tasks since) are using a multi $$ click-ed-y-click tool now. It works ...
I'm late, but who knows, a nice day you'll re-allocated for this task: for GraphViz, you can define not only the exact positions, but control rank levels. They're used, say, to keep date labels and events in the same column in a timeline graph.
There's a good example in this document, go to page 16: http://www.graphviz.org/pdf/dotguide.pdf
Anyway, if you're using ranking, you have to write the piece of code, which extracts ranking information from the source data, or, worst case, you have to add this information to items by hand.
