Graphhopper

PROD

Description: see README.

Usage

Create/Update S3 Bucket

The S3 bucket is managed via CFN Template in the .aws folder

aws cloudformation <create|update>-stack \
  --profile movelab \
  --stack-name lab-graphhopper-graphdata-prod \
  --template-body file://./.aws/graphdata.s3.yaml \
  --parameters ParameterKey=ServiceName,ParameterValue=lab-graphhopper-graphdata-prod

Build Image

docker build . -t docker.intapps.it/graphhopper:$(git rev-parse HEAD)

Start

docker run -p 8989:8989 docker.intapps.it/graphhopper:$(git rev-parse HEAD)

Merge PBF files via osmosis

osmosis \
  --read-pbf file=hamburg-latest.osm.pbf \
  --read-pbf file=berlin-latest.osm.pbf \
  --merge \
  --write-pbf \
  omitmetadata=true \
  file=b_hh_lndn_merged.pbf

Build + Deploy Graph

Start graphhopper, mount volume and pass path to pbf file.

NOTE: the -i expects the path from within the docker container context

docker run -p 8999:8989 -v </path/to/pbf/file/>:/data docker.intapps.it/graphhopper:$(git rev-parse HEAD) -i <path/to/pbf/file>.pbf

e.g.:

docker run -p 8999:8989 -v /Users/macuser/Desktop/pbf:/data docker.intapps.it/graphhopper:$(git rev-parse HEAD) -i /data/b_hh_lndn_dnmrk_merged.pbf

Zip the created pbf file

tar -zcvf <filename>.gz <path/to/build/graphdata>

Upload .gz file into the s3 bucket via

aws s3 cp /path/to/file.name s3://lab-graphhopper-graphdata-prod --profile movelab --region eu-west-1 --recursive

NOTE: If you want to use the compiled graph files use the following commands:

e.g.

docker run <...> -pg /data/graphdata/graphdata_b_hh_lndn.osm-gh -dp s3://lab-graphhopper-graphdata-prod/graphdata_b_hh_lndn.gz

The script will download the data from s3://lab-graphhopper-graphdata-prod/graphdata_b_hh_lndn.gz. Then graphdata_b_hh_lndn.gz will be extracted into a folder called graphdata_b_hh_lndn.osm-gh (In this case).

During the startup, graphhopper will look in /data/graphdata/graphdata_b_hh_lndn.osm-gh/ where the built graph is then located.