Your application can make two DB connections to gather that data from both side then manually process and sort it. Depending on what you are trying to do this can be a complete pain in the butt and very slow.
If you want it to be seamless you don't have a lot of choice but to use DB links, but you don't need to use them directly. Rather than having the data distributed you could consider replicating it. Materialized views (snapshots) allow one-way transfer of data, or two-way if the updatable option is used. If you wanted to you could use advanced replication but this increases the complexity drastically.
Tim gave you a very good suggestion. Yes, use the DB links but transfer the data via MVs. This will be the quickest way. Build the MV logs on the first node (this is a quick process). Build then the MVs on the second node. Then distribute it the way you want.
Oracle Certified Master
Oracle Certified Professional 6i,8i,9i,10g,11g