Running Multiple Processing Piplines

[Christopher Miles]

Hello,

If I want to have multiple processing pipelines (perhaps one for each type of document collection), it looks like the most straightforward solution is to simply run multiple instances of Pypes, each with the one pipeline of operations. Is this the recommended way of getting multiple processing pipelines? I want to make sure I'm not missing something. 

Thank you

--

Miles

[Eric Gaumer]

Yes that's correct. Being graph oriented, you can also create different branches of logic within the same instance. You can then write a "routing" stage that sends documents down specific paths based on some routing key (e.g., a field identifying the collection type).

Both approaches should work fine.

-Eric