1 # Importing the Database
3 The following instructions explain how to create a Nominatim database
4 from an OSM planet file. It is assumed that you have already successfully
5 installed the Nominatim software itself and the `nominatim` tool can be found
6 in your `PATH`. If this is not the case, return to the
7 [installation page](Installation.md).
9 ## Creating the project directory
11 Before you start the import, you should create a project directory for your
12 new database installation. This directory receives all data that is related
13 to a single Nominatim setup: configuration, extra data, etc. Create a project
14 directory apart from the Nominatim software and change into the directory:
17 mkdir ~/nominatim-planet
21 In the following, we refer to the project directory as `$PROJECT_DIR`. To be
22 able to copy&paste instructions, you can export the appropriate variable:
25 export PROJECT_DIR=~/nominatim-planet
28 The Nominatim tool assumes per default that the current working directory is
29 the project directory but you may explicitly state a different directory using
30 the `--project-dir` parameter. The following instructions assume that you run
31 all commands from the project directory.
33 !!! tip "Migration Tip"
35 Nominatim used to be run directly from the build directory until version 3.6.
36 Essentially, the build directory functioned as the project directory
37 for the database installation. This setup still works and can be useful for
38 development purposes. It is not recommended anymore for production setups.
39 Create a project directory that is separate from the Nominatim software.
41 ### Configuration setup in `.env`
43 The Nominatim server can be customized via a `.env` in the project directory.
44 This is a file in [dotenv](https://github.com/theskumar/python-dotenv) format
45 which looks the same as variable settings in a standard shell environment.
46 You can also set the same configuration via environment variables. All
47 settings have a `NOMINATIM_` prefix to avoid conflicts with other environment
50 There are lots of configuration settings you can tweak. Have a look
51 at `settings/env.default` for a full list. Most should have a sensible default.
55 If you plan to import a large dataset (e.g. Europe, North America, planet),
56 you should also enable flatnode storage of node locations. With this
57 setting enabled, node coordinates are stored in a simple file instead
58 of the database. This will save you import time and disk storage.
61 NOMINATIM_FLATNODE_FILE="/path/to/flatnode.file"
63 Replace the second part with a suitable path on your system and make sure
64 the directory exists. There should be at least 75GB of free space.
66 ## Downloading additional data
68 ### Wikipedia/Wikidata rankings
70 Wikipedia can be used as an optional auxiliary data source to help indicate
71 the importance of OSM features. Nominatim will work without this information
72 but it will improve the quality of the results if this is installed.
73 This data is available as a binary download. Put it into your project directory:
76 wget https://www.nominatim.org/data/wikimedia-importance.sql.gz
78 The file is about 400MB and adds around 4GB to the Nominatim database.
81 If you forgot to download the wikipedia rankings, you can also add
82 importances after the import. Download the files, then run
83 `nominatim refresh --wiki-data --importance`. Updating importances for
84 a planet can take a couple of hours.
86 ### Great Britain, USA postcodes
88 Nominatim can use postcodes from an external source to improve searches that
89 involve a GB or US postcode. This data can be optionally downloaded into the
93 wget https://www.nominatim.org/data/gb_postcode_data.sql.gz
94 wget https://www.nominatim.org/data/us_postcode_data.sql.gz
96 ## Choosing the data to import
98 In its default setup Nominatim is configured to import the full OSM data
99 set for the entire planet. Such a setup requires a powerful machine with
100 at least 64GB of RAM and around 900GB of SSD hard disks. Depending on your
101 use case there are various ways to reduce the amount of data imported. This
102 section discusses these methods. They can also be combined.
106 If you only need geocoding for a smaller region, then precomputed OSM extracts
107 are a good way to reduce the database size and import time.
108 [Geofabrik](https://download.geofabrik.de) offers extracts for most countries.
109 They even have daily updates which can be used with the update process described
110 [in the next section](../Update). There are also
111 [other providers for extracts](https://wiki.openstreetmap.org/wiki/Planet.osm#Downloading).
113 Please be aware that some extracts are not cut exactly along the country
114 boundaries. As a result some parts of the boundary may be missing which means
115 that Nominatim cannot compute the areas for some administrative areas.
117 ### Dropping Data Required for Dynamic Updates
119 About half of the data in Nominatim's database is not really used for serving
120 the API. It is only there to allow the data to be updated from the latest
121 changes from OSM. For many uses these dynamic updates are not really required.
122 If you don't plan to apply updates, you can run the import with the
123 `--no-updates` parameter. This will drop the dynamic part of the database as
124 soon as it is not required anymore.
126 You can also drop the dynamic part later using the following command:
132 Note that you still need to provide for sufficient disk space for the initial
133 import. So this option is particularly interesting if you plan to transfer the
134 database or reuse the space later.
136 ### Reverse-only Imports
138 If you only want to use the Nominatim database for reverse lookups or
139 if you plan to use the installation only for exports to a
140 [photon](https://photon.komoot.de/) database, then you can set up a database
141 without search indexes. Add `--reverse-only` to your setup command above.
143 This saves about 5% of disk space.
145 ### Filtering Imported Data
147 Nominatim normally sets up a full search database containing administrative
148 boundaries, places, streets, addresses and POI data. There are also other
149 import styles available which only read selected data:
151 * **settings/import-admin.style**
152 Only import administrative boundaries and places.
153 * **settings/import-street.style**
154 Like the admin style but also adds streets.
155 * **settings/import-address.style**
156 Import all data necessary to compute addresses down to house number level.
157 * **settings/import-full.style**
158 Default style that also includes points of interest.
159 * **settings/import-extratags.style**
160 Like the full style but also adds most of the OSM tags into the extratags
163 The style can be changed with the configuration `NOMINATIM_IMPORT_STYLE`.
165 To give you an idea of the impact of using the different styles, the table
166 below gives rough estimates of the final database size after import of a
167 2020 planet and after using the `--drop` option. It also shows the time
168 needed for the import on a machine with 64GB RAM, 4 CPUS and NVME disks.
169 Note that the given sizes are just an estimate meant for comparison of
170 style requirements. Your planet import is likely to be larger as the
171 OSM data grows with time.
173 style | Import time | DB size | after drop
174 ----------|--------------|------------|------------
175 admin | 4h | 215 GB | 20 GB
176 street | 22h | 440 GB | 185 GB
177 address | 36h | 545 GB | 260 GB
178 full | 54h | 640 GB | 330 GB
179 extratags | 54h | 650 GB | 340 GB
181 You can also customize the styles further.
182 A [description of the style format](../develop/Import.md#configuring-the-import)
183 can be found in the development section.
185 ## Initial import of the data
187 !!! danger "Important"
188 First try the import with a small extract, for example from
189 [Geofabrik](https://download.geofabrik.de).
191 Download the data to import. Then issue the following command
192 from the **build directory** to start the import:
195 nominatim import --osm-file <data file> 2>&1 | tee setup.log
198 ### Notes on full planet imports
200 Even on a perfectly configured machine
201 the import of a full planet takes around 2 days. Once you see messages
202 with `Rank .. ETA` appear, the indexing process has started. This part takes
203 the most time. There are 30 ranks to process. Rank 26 and 30 are the most complex.
204 They take each about a third of the total import time. If you have not reached
205 rank 26 after two days of import, it is worth revisiting your system
206 configuration as it may not be optimal for the import.
208 ### Notes on memory usage
210 In the first step of the import Nominatim uses [osm2pgsql](https://osm2pgsql.org)
211 to load the OSM data into the PostgreSQL database. This step is very demanding
212 in terms of RAM usage. osm2pgsql and PostgreSQL are running in parallel at
213 this point. PostgreSQL blocks at least the part of RAM that has been configured
214 with the `shared_buffers` parameter during
215 [PostgreSQL tuning](Installation#postgresql-tuning)
216 and needs some memory on top of that. osm2pgsql needs at least 2GB of RAM for
217 its internal data structures, potentially more when it has to process very large
218 relations. In addition it needs to maintain a cache for node locations. The size
219 of this cache can be configured with the parameter `--osm2pgsql-cache`.
221 When importing with a flatnode file, it is best to disable the node cache
222 completely and leave the memory for the flatnode file. Nominatim will do this
223 by default, so you do not need to configure anything in this case.
225 For imports without a flatnode file, set `--osm2pgsql-cache` approximately to
226 the size of the OSM pbf file you are importing. The size needs to be given in
227 MB. Make sure you leave enough RAM for PostgreSQL and osm2pgsql as mentioned
228 above. If the system starts swapping or you are getting out-of-memory errors,
229 reduce the cache size or even consider using a flatnode file.
232 ### Testing the installation
234 Run this script to verify all required tables and indices got created successfully.
237 nominatim admin --check-database
240 Now you can try out your installation by running:
246 This runs a small test server normally used for development. You can use it
247 to verify that your installation is working. Go to
248 `http://localhost:8088/status.php` and you should see the message `OK`.
249 You can also run a search query, e.g. `http://localhost:8088/search.php?q=Berlin`.
251 To run Nominatim via webservers like Apache or nginx, please read the
252 [Deployment chapter](Deployment.md).
254 ## Tuning the database
256 Accurate word frequency information for search terms helps PostgreSQL's query
257 planner to make the right decisions. Recomputing them can improve the performance
258 of forward geocoding in particular under high load. To recompute word counts run:
261 nominatim refresh --word-counts
264 This will take a couple of hours for a full planet installation. You can
265 also defer that step to a later point in time when you realise that
266 performance becomes an issue. Just make sure that updates are stopped before
267 running this function.
269 If you want to be able to search for places by their type through
270 [special key phrases](https://wiki.openstreetmap.org/wiki/Nominatim/Special_Phrases)
271 you also need to enable these key phrases like this:
273 nominatim special-phrases --from-wiki > specialphrases.sql
274 psql -d nominatim -f specialphrases.sql
276 Note that this command downloads the phrases from the wiki link above. You
277 need internet access for the step.
280 ## Installing Tiger housenumber data for the US
282 Nominatim is able to use the official [TIGER](https://www.census.gov/geographies/mapping-files/time-series/geo/tiger-line-file.html)
283 address set to complement the OSM house number data in the US. You can add
284 TIGER data to your own Nominatim instance by following these steps. The
285 entire US adds about 10GB to your database.
287 1. Get preprocessed TIGER 2020 data and unpack it into your project
291 wget https://nominatim.org/data/tiger2020-nominatim-preprocessed.tar.gz
292 tar xf tiger2020-nominatim-preprocessed.tar.gz
294 2. Import the data into your Nominatim database:
296 nominatim add-data --tiger-data tiger
298 3. Enable use of the Tiger data in your `.env` by adding:
300 echo NOMINATIM_USE_US_TIGER_DATA=yes >> .env
302 4. Apply the new settings:
304 nominatim refresh --functions
307 See the [developer's guide](../develop/data-sources.md#us-census-tiger) for more
308 information on how the data got preprocessed.