Integrating Geonames data into your Mint installation is reasonably simple, but can be a little time consuming. The Solr index is simply too large to distribute reasonably over the Internet at this stage, so at least the first time you do this, it will have to be built locally for your institution. First time buildIf you don't have a copy of a Geonames index sitting around to copy in place, you are going to have to build it. Start by downloading and unpacking the Geonames data dump (around 200mb). On *nix you can do this from the command line: wget http://download.geonames.org/export/dump/allCountries.zipunzip allCountries.zip
This will leave you with about 900mb of data in a text file (allCountries.txt) to ingest. You'll need to run the harvest script in the 'server' folder to ingest this: ./geo_harvest.sh allCountries.txt
PLEASE NOTE: At time of release (v1.1) *nix based systems don't having the correct logging configuration for this script. It will pour enormous amounts of data into 'main.log' (~1.5gb). This will be resolved ASAP, but in the interim it is recommended that you a) run this script whilst the server is offline, and b) backup and restore you 'main.log' before and after. The log entries themselves are not significant and do not need to be kept. During the v1.2 testing this problem has been improved somewhat. The excessive logging is still occurring, but it is confined to the correct log file and is trivial to remove afterwords. It is worth noting however some testing was performed on running this script with the server online. It is perfectly viable, but for now will require a server reboot upon completion to find the Solr index changes. If you want to watch the log for only important rows whilst the harvest is ongoing, try something like this: tail -f geo_harvest.out | grep read2012-04-13 12:35:28,790 INFO Harvester 12:35:28: Rows read: 600002012-04-13 12:35:44,105 INFO Harvester 12:35:44: Rows read: 80000System Re-Build
|


