1OSM2PGSQL(1) General Commands Manual OSM2PGSQL(1)
2
6 osm2pgsql - Openstreetmap data to PostgreSQL converter.
7
9 osm2pgsql [options] planet.osm
10 osm2pgsql [options] planet.osm.{gz,bz2,pbf}
11 osm2pgsql [options] file1.osm file2.osm file3.osm
12
14 This manual page documents briefly the osm2pgsql command.
15 osm2pgsql imports data from OSM file(s) into a PostgreSQL database
17 suitable for use by the Mapnik renderer or the Nominatim geocoder.
18 OSM planet snapshots can be downloaded from https://planet.open‐
19 streetmap.org/. Partial planet files ("extracts") for various countries
20 are available, see https://wiki.openstreetmap.org/wiki/Planet.osm.
21 Extracts in PBF (ProtoBufBinary) format are also available from
23 https://download.geofabrik.de/osm/.
24 When operating in "slim" mode (and on a database created in "slim"
26 mode!), osm2pgsql can also process OSM change files (osc files),
27 thereby bringing an existing database up to date.
28
30 These programs follow the usual GNU command line syntax, with long
31 options starting with two dashes (`-'). A summary of options is
32 included below.
33 -a|--append
35 Add the OSM file into the database without removing existing
36 data.
37 -b|--bbox
39 Apply a bounding box filter on the imported data. Must be spec‐
40 ified as: minlon,minlat,maxlon,maxlat e.g. --bbox
41 -0.5,51.25,0.5,51.75
42 -c|--create
44 Remove existing data from the database. This is the default if
45 --append is not specified.
46 -d|--database name
48 The name of the PostgreSQL database to connect to.
49 -i|--tablespace-index tablespacename
51 Store all indices in a separate PostgreSQL tablespace named by
52 this parameter. This allows one to e.g. store the indices on
53 faster storage like SSDs.
54 --tablespace-main-data tablespacename
56 Store the data tables (non slim) in the given tablespace.
57 --tablespace-main-index tablespacename
59 Store the indices of the main tables (non slim) in the given
60 tablespace.
61 --tablespace-slim-data tablespacename
63 Store the slim mode tables in the given tablespace.
64 --tablespace-slim-index tablespacename
66 Store the indices of the slim mode tables in the given
67 tablespace.
68 -l|--latlong
70 Store data in degrees of latitude & longitude.
71 -m|--merc
73 Store data in proper spherical Mercator (the default).
74 -E|--proj num
76 Use projection EPSG:num
77 -p|--prefix prefix_string
79 Prefix for table names (default: planet_osm).
80 -r|--input-reader format
82 Select format of the input file. Available choices are auto
83 (default) for autodetecting the format, xml for OSM XML format
84 files, o5m for o5m formatted files and pbf for OSM PBF binary
85 format.
86 -s|--slim
88 Store temporary data in the database. Without this mode, all
89 temporary data is stored in RAM and if you do not have enough
90 the import will not work successfully. With slim mode, you
91 should be able to import the data even on a system with limited
92 RAM, although if you do not have enough RAM to cache at least
93 all of the nodes, the time to import the data will likely be
94 greatly increased.
95 --drop
97 Drop the slim mode tables from the database once the import is
98 complete. This can greatly reduce the size of the database, as
99 the slim mode tables typically are the same size, if not
100 slightly bigger than the main tables. It does not, however,
101 reduce the maximum spike of disk usage during import. It can
102 furthermore increase the import speed, as no indices need to be
103 created for the slim mode tables, which (depending on hardware)
104 can nearly halve import time. Slim mode tables however have to
105 be persistent if you want to be able to update your database, as
106 these tables are needed for diff processing.
107 -S|--style /path/to/style
109 Location of the osm2pgsql style file. This specifies which tags
110 from the data get imported into database columns and which tags
111 get dropped. Defaults to /usr/share/osm2pgsql/default.style.
112 -C|--cache num
114 Only for slim mode: Use up to num many MB of RAM for caching
115 nodes. Giving osm2pgsql sufficient cache to store all imported
116 nodes typically greatly increases the speed of the import. Each
117 cached node requires 8 bytes of cache, plus about 10% - 30%
118 overhead. For a current OSM full planet import with its ~ 3 bil‐
119 lion nodes, a good value would be 27000 if you have enough RAM.
120 If you don't have enough RAM, it is likely beneficial to give
121 osm2pgsql close to the full available amount of RAM. Defaults to
122 800.
123 --cache-strategy strategy
125 There are a number of different modes in which osm2pgsql can
126 organize its node cache in RAM. These are optimized for differ‐
127 ent assumptions of the data and the hardware resources avail‐
128 able. Currently available strategies are dense, chunked, sparse
129 and optimized. dense assumes that the node id numbers are
130 densely packed, i.e. only a few IDs in the range are missing /
131 deleted. For planet extracts this is usually not the case, mak‐
132 ing the cache very inefficient and wasteful of RAM. sparse
133 assumes node IDs in the data are not densely packed, greatly
134 increasing caching efficiency in these cases. If node IDs are
135 densely packed, like in the full planet, this strategy has a
136 higher overhead for indexing the cache. optimized uses both
137 dense and sparse strategies for different ranges of the ID
138 space. On a block by block basis it tries to determine if it is
139 more effective to store the block of IDs in sparse or dense
140 mode. This is the default and should be typically used.
141 -U|--username name
143 Postgresql user name.
144 -W|--password
146 Force password prompt.
147 -H|--host hostname
149 Database server hostname or socket location.
150 -P|--port num
152 Database server port.
153 -e|--expire-tiles [min_zoom-]max-zoom
155 Create a tile expiry list.
156 -o|--expire-output /path/to/expire.list
158 Output file name for expired tiles list.
159 -O|--output
161 Specifies the output back-end or database schema to use. Cur‐
162 rently osm2pgsql supports pgsql, gazetteer and null. pgsql is
163 the default output back-end / schema and is optimized for ren‐
164 dering with Mapnik. gazetteer is a db schema optimized for
165 geocoding and is used by Nominatim. null does not write any
166 output and is only useful for testing or with --slim for creat‐
167 ing slim tables.
168 -x|--extra-attributes
170 Include attributes for each object in the database. This
171 includes the username, userid, timestamp and version. Note:
172 this option also requires additional entries in your style file.
173 -k|--hstore
175 Add tags without column to an additional hstore (key/value) col‐
176 umn to PostgreSQL tables.
177 -j|--hstore-all
179 Add all tags to an additional hstore (key/value) column in Post‐
180 greSQL tables.
181 -z|--hstore-column key_name
183 Add an additional hstore (key/value) column containing all tags
184 that start with the specified string, eg --hstore-column "name:"
185 will produce an extra hstore column that contains all name:xx
186 tags
187 --hstore-match-only
189 Only keep objects that have a value in one of the columns (nor‐
190 mal action with --hstore is to keep all objects).
191 --hstore-add-index
193 Create indices for the hstore columns during import.
194 -G|--multi-geometry
196 Normally osm2pgsql splits multi-part geometries into separate
197 database rows per part. A single OSM id can therefore have sev‐
198 eral rows. With this option, PostgreSQL instead generates
199 multi-geometry features in the PostgreSQL tables.
200 -K|--keep-coastlines
202 Keep coastline data rather than filtering it out. By default
203 natural=coastline tagged data will be discarded based on the
204 assumption that post-processed Coastline Checker shape files
205 will be used.
206 --exclude-invalid-polygon
208 OpenStreetMap data is defined in terms of nodes, ways and rela‐
209 tions and not in terms of actual geometric features. Osm2pgsql
210 therefore tries to build postgis geometries out of this data
211 representation. However not all ways and relations correspond to
212 valid PostGIS geometries (e.g. self intersecting polygons). By
213 default osm2pgsql tries to fix these geometries using buffer(0)
214 around the invalid polygons. With this option, invalid polygons
215 are instead simply dropped from the database. Even without this
216 option, all polygons in the database should be valid.
217 --unlogged
219 Use postgresql's unlogged tables for storing data. This requires
220 PostgreSQL 9.1 or above. Data written to unlogged tables is not
221 written to PostgreSQL's write-ahead log, which makes them con‐
222 siderably faster than ordinary tables. However, they are not
223 crash-safe: an unlogged table is automatically truncated after a
224 crash or unclean shutdown.
225 --number-processes num
227 Specifies the number of parallel processes used for certain
228 operations. If disks are fast enough e.g. if you have an SSD,
229 then this can greatly increase speed of the "going over pending
230 ways" and "going over pending relations" stages on a multi-core
231 server.
232 -I|--disable-parallel-indexing
234 By default osm2pgsql initiates the index building on all tables
235 in parallel to increase performance. This can be disadvantages
236 on slow disks, or if you don't have enough RAM for PostgreSQL to
237 perform up to 7 parallel index building processes (e.g. because
238 maintenance_work_mem is set high).
239 --flat-nodes /path/to/nodes.cache
241 The flat-nodes mode is a separate method to store slim mode node
242 information on disk. Instead of storing this information in the
243 main PostgreSQL database, this mode creates its own separate
244 custom database to store the information. As this custom data‐
245 base has application level knowledge about the data to store and
246 is not general purpose, it can store the data much more effi‐
247 ciently. Storing the node information for the full planet
248 requires about 100GB in PostgreSQL, the same data is stored in
249 only ~16GB using the flat-nodes mode. This can also increase the
250 speed of applying diff files. This option activates the
251 flat-nodes mode and specifies the location of the database file.
252 It is a single large > 16GB file. This mode is only recommended
253 for full planet imports as it doesn't work well with small
254 imports. The default is disabled.
255 -h|--help
257 Help information.
258 Add -v to display supported projections.
259 -v|--verbose
261 Verbose output.
262
264 Latlong (-l) SRS: 4326 (none)
265 Spherical Mercator (-m) SRS:3857 +proj=merc +a=6378137 +b=6378137
266 +lat_ts=0.0 +lon_0=0.0 +x_0=0.0 +y_0=0 +k=1.0 +units=m +nadgrids=@null
267 +no_defs +over
268 EPSG-defined (-E) SRS: +init=epsg:(as given in parameter)
269
271 proj(1), postgres(1).
272
274 osm2pgsql was written by Jon Burgess, Artem Pavlenko, and other Open‐
275 StreetMap project members.
276 This manual page was written by Andreas Putzo <andreas@putzo.net> for
278 the Debian project, and amended by OpenStreetMap authors.
279 February 5, 2017 OSM2PGSQL(1)