Hi, all!
I've havent found an explicit answer to this:
If I create N publishers on one topic, each publisher publishing data for one particular topic key (so we have N topic instances), and then one subscriber to the topic that wants to get only data for one particular key. Is the data already filtered (by key) on the publisher side or the subscriber gets all the data (and filters locally)?
Thank you,
Matej
Hi Matej,
The answer is "it depends". If you create a content filtered topic datareader with a filter string that accepts only the one key, then the filtering /may/ take place on the writer side (writer-side filtering). There are some QoS settings that will prevent writer-side, however, and so filtering will happen at the reader. Search for "writer-side filtering" in the PDF doc.
The thing to remember is that with a CFT, your application will only get the instances that pass the filter -- irrespective of where the filtering took place.
additional info: CFT don't need to filter on the key, they can filter on any field (when using the SQL-like where clause), or can filter even on complex algorithm if using a custom content filter.
Regards,
rip