Merge pull request #240 from CartoDB/231-Update_IP_address_database_for_geocoding
IP geocoder table update
This commit is contained in:
@@ -11,12 +11,61 @@ SELECT geocode_ip(Array['1.0.16.0', '::ffff:1.0.16.0'])
|
||||
```
|
||||
|
||||
# Creation steps
|
||||
1. Create the `ip_address_locations` table
|
||||
1. Create the `ip_address_locations` table (see `40_ipaddr.sql` file)
|
||||
2. Obtain the file from http://geolite.maxmind.com/download/geoip/database/GeoLite2-City-CSV.zip
|
||||
3. Uncompress it and upload the `GeoLite2-City-Blocks-IPv4.csv` file
|
||||
4. Rename the uploaded table as `latest_ip_address_locations`
|
||||
5. Run the `sql/build_data_table` script to update the table
|
||||
|
||||
# Update steps
|
||||
|
||||
## Option A: generate a new `ip_address_locations` table at geocoder user
|
||||
|
||||
If the geocoder database is a CARTO user, do these steps:
|
||||
|
||||
1. Import `GeoLite2-City-Blocks-IPv4.csv` and rename it to `latest_ip_address_locations`.
|
||||
2. Import `GeoLite2-City-Blocks-IPv6.csv` and rename it to `latest_ip6_address_locations`.
|
||||
3. If you want to create a backup of the previous table, do this:
|
||||
|
||||
```sql
|
||||
CREATE TABLE ip_address_locations_backup as
|
||||
select * from ip_address_locations;
|
||||
```
|
||||
|
||||
4. Clear previous table:
|
||||
|
||||
```sql
|
||||
TRUNCATE ip_address_locations;
|
||||
```
|
||||
|
||||
5. Load the new values:
|
||||
|
||||
```sql
|
||||
set statement_timeout = '20min';
|
||||
INSERT INTO ip_address_locations (the_geom, network_start_ip) SELECT the_geom, ('::ffff:' || split_part(network, '/', 1))::inet FROM latest_ip_address_locations;
|
||||
INSERT INTO ip_address_locations (the_geom, network_start_ip) SELECT the_geom, split_part(network, '/', 1)::inet FROM latest_ip6_address_locations;
|
||||
```
|
||||
|
||||
## Option B: load a dump of the table
|
||||
|
||||
If the geocoder database is not a CARTO user, do these steps:
|
||||
|
||||
1. Perform option A in any user (it can even be a staging user). If you need to create the table, check `40_ipaddr.sql`.
|
||||
2. Generate a dump of the file at that user database:
|
||||
|
||||
```sql
|
||||
\copy (select * from ip_address_locations) TO /tmp/ip_address_locations.dump;
|
||||
```
|
||||
|
||||
3. Copy the file to the remote server.
|
||||
4. Perform steps A.3 and A.4.
|
||||
5. Load the new data (takes ~10 minutes):
|
||||
|
||||
```sql
|
||||
set statement_timeout = '20min';
|
||||
\copy ip_address_locations from /tmp/ip_address_locations.dump
|
||||
```
|
||||
|
||||
# Tables
|
||||
|
||||
### ip_address_locations
|
||||
|
||||
@@ -4,5 +4,5 @@
|
||||
-- Clear table
|
||||
DELETE FROM ip_address_locations;
|
||||
-- Updates table with new source data
|
||||
INSERT INTO ip_address_locations (the_geom, network_start_ip) SELECT the_geom, network_start_ip::inet FROM latest_ip_address_locations;
|
||||
INSERT INTO ip_address_locations (the_geom, network_start_ip) SELECT the_geom, ('::ffff:' || split_part(network, '/', 1))::inet FROM latest_ip_address_locations;
|
||||
DROP TABLE latest_ip_address_locations;
|
||||
|
||||
Reference in New Issue
Block a user