Hi Carlos, I've tried with 3'' files and got this SCHWERWIEGEND (MapFailedException): split\63240001.osm.pbf: (thrown in BufferedImgFileWriter.ensureSize()) There is not enough room in a single garmin map for all the input data. The .osm file should be split into smaller pieces first. SCHWERWIEGEND (MapBuilder): split\63240001.osm.pbf: exception while creating DEM file There is not enough room in a single garmin map for all the input data. The .osm file should be split into smaller pieces first. SCHWERWIEGEND (MapFailedException): split\63240001.osm.pbf: (thrown in MapBuilder.makeMap()) DEM I think the reason is that you use --block-size=2048. Please remove the option so that mkgmap can calculate the best value. The traceback shows that the crash happens while mkgmap calculates DEM for the overview map. I wonder why it has an effect when you split the tiles. This should not happen. I cannot reproduce this problem so far. Maybe you have a special *.poly file? I used http://download.geofabrik.de/africa/algeria.poly Please post your file and the original areas.list produced by splitter. Besides that I can reproduce the performance problems, you should either limit the number of parallel jobs with the max-jobs value or provide more heap with -Xmx option or reduce splitter option max-nodes so that you don't create such large DEM files. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Carlos Dávila <cdavilam@orangecorreo.es> Gesendet: Dienstag, 27. Februar 2018 23:06:00 An: Development list for mkgmap Betreff: [mkgmap-dev] Error building DEM map I'm facing an error compiling Algeria map with DEM, using SRTM1 hgt files. In a first run I use splitter as follows: java -Xmx6G -jar splitter.jar --max-nodes=2000000 --geonames-file=geonames/cities15000_$ABR.zip --mapid=551${FID}001 --output=o5m algeria.o5m It gives 5 tiles and when I run mkgmap on them [1], it takes a very long time and finally it fails with "Number of MapFailedExceptions: 1" and a huge (+7GB) log file I could not read. I have manually split problematic tile to a half its original size and then both resulting tiles fail to compile. After dividing them /2, 3 of the 4 tiles fail, but then mkgmap gives the following error: Exception in thread "main" java.lang.NegativeArraySizeException at uk.me.parabola.mkgmap.reader.hgt.HGTConverter.<init>(HGTConverter.java:85) at uk.me.parabola.imgfmt.app.dem.DEMFile.calc(DEMFile.java:76) at uk.me.parabola.mkgmap.build.MapBuilder.makeMap(MapBuilder.java:341) at uk.me.parabola.mkgmap.combiners.OverviewBuilder.writeOverviewMap(OverviewBuilder.java:191) at uk.me.parabola.mkgmap.combiners.OverviewBuilder.onFinish(OverviewBuilder.java:104) at uk.me.parabola.mkgmap.main.Main.endOptions(Main.java:679) at uk.me.parabola.mkgmap.CommandArgsReader.readArgs(CommandArgsReader.java:128) at uk.me.parabola.mkgmap.main.Main.mainStart(Main.java:143) at uk.me.parabola.mkgmap.main.Main.main(Main.java:114) And log contains over 10⁷ lines like this: WARNING (HGTConverter): 55129161.o5m: height interpolation returns void at 30.837045,2.094557 After 2 more /2 divisions I could build all tiles. The smallest failing tile can be downloaded from http://alternativaslibres.org/tmp/55129083.o5m [1] java -Xmx4500m -ea -Dlog.config=logging.properties -jar mkgmap.jar --block-size=2048 --dem=hgt/VIEW1,hgt/SRTM1,hgt/EU-DEM,hgt/VIEW3,hgt --dem-poly=polygons/$pais.poly --dem-dists=3312,6624,9936,13248,26512 --overview-dem-dist=128000 --output-dir=./tmp --gmapi --show-profiles=1 --bounds=bounds.zip --precomp-sea=sea.zip --max-jobs --route --latin1 --code-page=1252 --country-name=$PAIS --country-abbr=$ABR --area-name=$MAPA --family-name="OSM $MAPA DEM" --family-id=5$FID --product-id=1 --product-version=$VERSION --series-name="OSM-$MAPA-DEM" --overview-mapname=$ABR-5$FID --overview-mapnumber=555${FID}000 --name-tag-list=$NAMETAG --index --process-destination --process-exits --housenumbers --reduce-point-density=4 --polygon-size-limits="24:12, 18:10, 16:0" --add-pois-to-areas --link-pois-to-ways --location-autofill=is_in,nearest --drive-on=detect,$DRIVEON --check-roundabouts --check-roundabout-flares --license-file=license_OSM_${HGT} --copyright-file=copyright $LANGUAGE --road-name-config=$CONFIG --x-mdr7-del=GRADE0,GRADE1,GRADE2,GRADE3,GRADE4,GRADE5,GRADE6,GRADE7,UNNAMED --style=mio --remove-ovm-work-files=true -c ${pais}-DEM.args _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

Hi Carlos, I remember that mkgmap takes very long when there are some strange large tiles which are almost empty (or contain large almost empty areas). I could solve that by cutting the desired area from the input file before running splitter. I do not know if this is applicable in your situation, too. You might look into the "areas" file generated by splitter: if there is such a strange large area, try the trick. Regards, Bernhard Am 27.02.2018 um 23:06 schrieb Carlos Dávila:

I'm facing an error compiling Algeria map with DEM, using SRTM1 hgt files. In a first run I use splitter as follows: java -Xmx6G -jar splitter.jar --max-nodes 00000 --geonames-file=onames/cities15000_$ABR.zip --mapidU1${FID}001 --output=m algeria.o5m It gives 5 tiles and when I run mkgmap on them [1], it takes a very long time and finally it fails with "Number of MapFailedExceptions: 1" and a huge (+7GB) log file I could not read. I have manually split problematic tile to a half its original size and then both resulting tiles fail to compile. After dividing them /2, 3 of the 4 tiles fail, but then mkgmap gives the following error: Exception in thread "main" java.lang.NegativeArraySizeException at uk.me.parabola.mkgmap.reader.hgt.HGTConverter.<init>(HGTConverter.java:85) at uk.me.parabola.imgfmt.app.dem.DEMFile.calc(DEMFile.java:76) at uk.me.parabola.mkgmap.build.MapBuilder.makeMap(MapBuilder.java:341) at uk.me.parabola.mkgmap.combiners.OverviewBuilder.writeOverviewMap(OverviewBuilder.java:191) at uk.me.parabola.mkgmap.combiners.OverviewBuilder.onFinish(OverviewBuilder.java:104) at uk.me.parabola.mkgmap.main.Main.endOptions(Main.java:679) at uk.me.parabola.mkgmap.CommandArgsReader.readArgs(CommandArgsReader.java:128) at uk.me.parabola.mkgmap.main.Main.mainStart(Main.java:143) at uk.me.parabola.mkgmap.main.Main.main(Main.java:114) And log contains over 10⁷ lines like this: WARNING (HGTConverter): 55129161.o5m: height interpolation returns void at 30.837045,2.094557 After 2 more /2 divisions I could build all tiles. The smallest failing tile can be downloaded from http://alternativaslibres.org/tmp/55129083.o5m

[1] java -Xmx4500m -ea -Dlog.config=gging.properties -jar mkgmap.jar --block-size 48 --dem=hgt/VIEW1,hgt/SRTM1,hgt/EU-DEM,hgt/VIEW3,hgt --dem-poly=lygons/$pais.poly --dem-dists312,6624,9936,13248,26512 --overview-dem-dist8000 --output-dir=./tmp --gmapi --show-profiles=1 --bounds=unds.zip --precomp-sea=sea.zip --max-jobs --route --latin1 --code-page52 --country-name=$PAIS --country-abbr=$ABR --area-name=APA --family-name="OSM $MAPA DEM" --family-id=5$FID --product-id=--product-version=$VERSION --series-name="OSM-$MAPA-DEM" --overview-mapname=BR-5$FID --overview-mapnumberU5${FID}000 --name-tag-list=AMETAG --index --process-destination --process-exits --housenumbers --reduce-point-density=--polygon-size-limits="24:12, 18:10, 16:0" --add-pois-to-areas --link-pois-to-ways --location-autofill=_in,nearest --drive-on�tect,$DRIVEON --check-roundabouts --check-roundabout-flares --license-file=cense_OSM_${HGT} --copyright-file=copyright $LANGUAGE --road-name-config=ONFIG --x-mdr7-del=ADE0,GRADE1,GRADE2,GRADE3,GRADE4,GRADE5,GRADE6,GRADE7,UNNAMED --style=o --remove-ovm-work-files=true -c ${pais}-DEM.args