A hyper-flexible Docker image for the excellent Valhalla routing framework.
# download a file to custom_files and start valhalla mkdir custom_files wget -O custom_files/andorra-latest.osm.pbf https://download.geofabrik.de/europe/andorra-latest.osm.pbf docker run -dt --name valhalla_gis-ops -p 8002:8002 -v $PWD/custom_files:/custom_files ghcr.io/gis-ops/docker-valhalla/valhalla:latest # or let the container download the file for you docker run -dt --name valhalla_gis-ops -p 8002:8002 -v $PWD/custom_files:/custom_files -e tile_urls=https://download.geofabrik.de/europe/andorra-latest.osm.pbf ghcr.io/gis-ops/docker-valhalla/valhalla:latest
This image aims at being user-friendly and most efficient with your time and resources. Once built, you can easily change Valhalla's configuration, the underlying OSM data graphs are built from, accompanying data (like Admin Areas, elevation tiles) or even pre-built graph tiles. Upon docker restart <container>
those changes are taken into account via hashed files, and, if necessary, new graph tiles will be built automatically.
**new**: supports advanced user management to avoid sudo access to host-shared folders and files, see [notes on user management](#notes-on-user-management)
NOTE, with the recent (17.03.2023) announcement of Docker to remove free "teams" (even those providing FOSS like us), we moved our images to Github packages. If it's not on Github you'll find an image version still on Dockerhub.
Our package registry provides the following:
latest
, updated from Valhalla Github repository every Saturday morningNote, you might have to do a
docker login
before.
If you want to build the image yourself, be aware that you might need to adapt the base image in the Dockerfile
to reflect the version of Valhalla you'd like to build. You can find the tags of the ghcr.io/valhalla/valhalla
images here: https://github.com/valhalla/valhalla/pkgs/container/valhalla. On top of the Valhalla base image we support the following build arguments (see notes on user management):
VALHALLA_UID
: specify the user UID for the container-internal valhalla
user. Either leave this blank or, most usually, specify your current user's UID.VALHALLA_GID
: specify the group GID for the container-internal valhalla
user. Either leave this blank or, most usually, specify the group's GID whose members you want to have write access to the container-generated files.Note, before Valhalla version 3.1.0
the building scheme was completely different. Please contact enquiry@gis-ops.com if you need access to previous Valhalla versions via Docker.
Then it's a simple
docker build -t ghcr.io/gis-ops/docker-valhalla/valhalla:latest .
This image respects the following custom environment variables to be passed during container startup. Note, all variables have a default:
tile_urls
: Add as many (space-separated) URLs as you like, e.g. https://download.geofabrik.de/europe/andorra-latest.osm.pbfuse_tiles_ignore_pbf
: True
uses a local tile.tar file and skips building. Default True
.force_rebuild
: True
forces a rebuild of the routing tiles and sets build_tar
to Force
. Default False
.build_elevation
: True
downloads elevation tiles which are covering the routing graph. Force
will do the same, but first delete any existing elevation tiles. Default False
.build_admins
: True
builds the admin db needed for things like border-crossing penalties and detailed routing responses. Force
will do the same, but first delete the existing db. Default False
.build_time_zones
: True
builds the timezone db which is needed for time-dependent routing. Force
will do the same, but first delete the existing db. Default False
.build_transit
: True
will attempt to build transit tiles if none exist yet. Force
will remove existing transit and routing tiles. Default False
.build_tar
(since 29.10.2021/v3.1.5
): True
creates a tarball of the tiles including an index which allows for extremely faster graph loading after reboots. Force
will do the same, but first delete the existing tarball. Default True
.server_threads
: How many threads valhalla_build_tiles
will use and valhalla_service
will run with. Default is the value of nproc
. If valhalla kills it self when building tiles, lower this number.path_extension
: This path will be appended to the container-internal /custom_files
(and by extension to the docker volume mapped to that path) and will be the directory where all files will be created. Can be very useful in certain deployment scenarios. No leading/trailing path separator allowed. Default is ''.serve_tiles
: True
starts the valhalla service. Default True
.tileset_name
: The name of the resulting graph on disk. Very useful in case you want to build multiple datasets in the same directory. Default valhalla_tiles
.traffic_name
: The name of the traffic.tar. Again, useful for serving mulitple traffic archives from the same directory. If empty, i.e. "", then no traffic archive will be built. Default traffic.tar
.update_existing_config
: True
updates missing keys in existing valhalla.json. Useful for updating stale config files to include newly introduced config parameters. Default True
.use_default_speeds_config
: True
loads a JSON file used to enhance default speeds (or falls back to an existing custom_files/default_speeds.json
) and sets the respective config entry. Read more here. Default False
.default_speeds_config_url
: Remote location of the default_speeds_config
JSON. Default https://raw.githubusercontent.com/OpenStreetMapSpeeds/schema/master/default_speeds.json
For the following instructions to work, you'll need to have the image locally available already, either from Github Docker registry or from local.
Start a background container from that image:
docker run -dt -v $PWD/custom_files:/custom_files -p 8002:8002 --name valhalla ghcr.io/gis-ops/docker-valhalla/valhalla:latest
The important part here is, that you map a volume from your host machine to the container's /custom_files
. The container will dump all relevant Valhalla files to that directory.
At this point Valhalla is running, but there is no graph tiles yet. Follow the steps below to customize your Valhalla instance in a heartbeat.
Note, alternatively you could create
custom_files
on your host before starting the container with all necessary files you want to be respected, e.g. the OSM PBF files.
Valhalla supports reading raw GTFS feeds to build transit into its graph, see the docs for more details.
Put the unzipped GTFS feeds as subfolders in the main gtfs folder, e.g. gtfs_feeds/berlin/
, otherwise the files will not be found.
To enable multimodal
routing, you'll need to map the directory which contains all the GTFS feeds to the container's /gtfs_feeds
directory, e.g.
docker run -dt -v gtfs_feeds:/gtfs_feeds -v $PWD/custom_files:/custom_files -p 8002:8002 --name valhalla gisops/valhalla:latest
Just dump single or multiple OSM PBF files to your mapped custom_files
directory, restart the container and Valhalla will start building the graphs:
cd custom_files # Download Andorra & Faroe Islands wget http://download.geofabrik.de/europe/faroe-islands-latest.osm.pbf http://download.geofabrik.de/europe/andorra-latest.osm.pbf docker restart valhalla
If you change the PBF files by either adding new ones or deleting any, Valhalla will build new tiles on the next restart unless told not to (e.g. setting use_tiles_ignore_pbf=True
).
Elevation tiles need to be in HGT (file format of the SRTM dataset) format and need to be named like NXXEYYY.hgt
. More info about format.
You need to store elevation tiles in the <base_path>/elevation_data
directory (by default custom_files/elevation_data/
). Tiles need to be grouped in folders by latitude , for example:
custom_files/elevation_data/
N53/
N53E016.hgt
N53E017.hgt
N54/
N54E016.hgt
If you had an existing graph before you acquired new elevation data, you'll need to rebuild the graph for the new data to become available, e.g. by starting a new container with force_rebuild=True. If you want to use new elevation data for Elevation API, you just need to restart container.
If you need to customize Valhalla's configuration to e.g. increase the allowed maximum distance for the /route
POST endpoint, just edit custom_files/valhalla.json
and restart the container. It won't rebuild the tiles in this case, unless you tell it to do so via environment variables.
In the case where you have a pre-built valhalla_tiles.tar
package from another Valhalla instance, you can also dump that to /custom_files/
and they're loaded upon container restart if you set the following environment variables: use_tiles_ignore_pbf=True
, force_rebuild=False
. Also, don't forget to set the md5 sum for your valhalla_tiles.tar
in .file_hashes.txt
.
If you want to verify that the image is working correctly, there's a small test script in ./tests
. Note, it might require sudo
, since it touches a few things generated by the container's valhalla
user:
./tests/test.sh
This project was first introduced by MichaelsJP.
AI辅助编程,代码自动修复
Trae是一种自适应的集成开发环境(IDE),通过自动化和多元协作改变开发流程。利用Trae,团队能够更快速、精确地编写和部署代码,从而提高编程效率和项目交付速度。Trae具备上下文感知和代码自动完成功能,是提升开发效率的理想工具。
AI小说写作助手,一站式润色、改写、扩写
蛙蛙写作—国内先进的AI写作平台,涵盖小说、学术、社交媒体等多场景。提供续写、改写、润色等功能,助力创作者高效优化写作流程。界面简洁,功能全面,适合各类写作者提升内容品质和工作效率。
全能AI智能助手,随时解答生活与工作的多样问题
问小白,由元石科技研发的AI智能助手,快速准确地解答各种生活和工作问题,包括但不限于搜索、规划和社交互动,帮助用户在日常生活中提高效率,轻松管理个人事务。
实时语音翻译/同声传译工具
Transly是一个多场景的AI大语言模型驱动的同声传译、专业翻译助手,它拥有超精准的音频识别翻译能力,几乎零延迟的使用体验和支持多国语言可以让你带它走遍全球,无论你是留学生、商务人士、韩剧美剧爱好者,还是出国游玩、多国会议、跨国追星等等,都可以满足你所有需要同传的场景需求,线上线下通用,扫除语言障碍,让全世界的语言交流不再有国界。
一键生成PPT和Word,让学习生活更轻松
讯飞智文是一个利用 AI 技术的项目,能够帮助用户生成 PPT 以及各类文档。无论是商业领域的市场分析报告、年度目标制定,还是学生群体的职业生涯规划、实习避坑指南,亦或是活动策划、旅游攻略等内容,它都能提供支持,帮助用户精准表达,轻松呈现各种信息。
深度推理能力全新升级,全面对标OpenAI o1
科大讯飞的星火大模型,支持语言理解、知识问答和文本创作等多功能,适用于多种文件和业务场景,提升办公和日常生活的效率。讯飞星火是一个提供丰富智能服务的平台,涵盖科技资讯、图像创作、写作辅助、编程解答、科研文献解读等功能,能为不同需求的用户提供便捷高效的帮助,助力用户轻松获取信息、解决问题,满足多样化使用场景。
一种基于大语言模型的高效单流解耦语音令牌文本到语音合成模型
Spark-TTS 是一个基于 PyTorch 的开源文本到语音合成项目,由多个知名机构联合参与。该项目提供了高效的 LLM(大语言模型)驱动的语音合成方案,支持语音克隆和语音创建功能,可通过命令行界面(CLI)和 Web UI 两种方式使用。用户可以根据需求调整语音的性别、音高、速度等参数,生成高质量的语音。该项目适用于多种场景,如有声读物制作、智能语音助手开发等。
AI助力,做PPT更简单!
咔片是一款轻量化在线演示设计工具,借助 AI 技术,实现从内容生成到智能设计的一站式 PPT 制作服务。支持多种文档格式导入生成 PPT,提供海量模板、智能美化、素材替换等功能,适用于销售、教师、学生等各类人群,能高效制作出高品质 PPT,满足不同场景演示需求。
选题、配图、成文,一站式创作,让内容运营更高效
讯飞绘文,一个AI集成平台,支持写作、选题、配图、排版和发布。高效生成适用于各类媒体的定制内容,加速品牌传播,提升内容营销效果。
专业的AI公文写作平台,公文写作神器
AI 材料星,专业的 AI 公文写作辅助平台,为体制内工作人员提供高效的公文写作解决方案。拥有海量公文文库、9 大核心 AI 功能,支持 30 + 文稿类型生成,助力快速完成领导讲话、工作总结、述职报告等材料,提升办公效率,是体制打工人的得力写作神器。
最新AI工具、AI资讯
独家AI资源、AI项目落地
微 信扫一扫关注公众号