Help updating OSM data using flat nodes file

The "since you have to reimport..." quote above was in response to someone who had not used the --flatnodes option during import hence it doesn't apply to you.

Applying diff updates with or without flat nodes is the same procedure, just that one needs the --flatnodes option on initial import and update, the other not.

You need to download all the diffs since your planet file was created, and apply them. For best performance, combine the diffs into one before you run osm2pgsql (using e.g. osmosis). If you want to configure continuous updates, use the --rrii Osmosis option initially and then --rri for each update. If you don't like to take guesses about when to start your replication, you can find out the highest node ID in your flat node file (osm2pgsql, when started in append mode, will usually say "highest node id in flat nodes file is ..."), then use the https://svn.openstreetmap.org/applications/utils/whichdiff/ utility to find the matching state.txt.