View Single Post
OmniPlan's scripting support will probably handle this, but it hasn't really been tested in that type of scenario.

All the properties of tasks and resources are scriptable at this point. (Let me know if something isn't!) Exporting is scriptable; however filtering on export is not. Printing is not scriptable.

There are two approaches (and possibly a hybrid, depending on the relationship between your projects).

1) Each of your projects in a separate document. Tell OmniPlan "make new document" and it will create a new document with one project in it. This will probably give you better performance. However, it does mean you'll have 10,000 files on your server. Also, you won't be able to level resources across projects automatically. (You could set up a script to approximate leveling across projects -- let me know if you need details.)
If your projects are relatively independent, this is probably the way to go. If you need to export one project, without all the others, this is the way you must do it, since filtering on export is not scriptable. In this scenario, you should be careful to close documents when you're done with them. OmniPlan cannot handle having all 10,000 documents open at once.

2) All of your projects in one OmniPlan document. An OmniPlan document technically only has one project, but that project could have 10,000 top-level task groups, one for each of your projects. This would allow you to level resources across your entire system. This will probably be slower, though -- even to update one task, let alone making changes that affect the whole document. This is the way to go if 10,000 files is a problem, or if you must level across the whole system, and can't approximate it by leveling each project in succession, and subtracting its schedule from the next project.

3) If there are logical sub-groupings of your projects -- by geographic region, by manager, by start date, whatever makes sense for you -- you could use a hybrid approach, with 100 documents, each containing 100 of your projects as a top-level task group. This might give you the best of both worlds, or the worst, depending on the exact nature of your data.