No matter how hard I looked, I couldn't find a good way.
How to install gdal in aws lambda?
If my question is not enough, please correct me. sorry for my english.
A good option IMO is to use a lambda layer. You can either create one yourself or use one which has been compiled and made available publicly (availability depends on your region).
Here are two options:
https://github.com/lambgeo/docker-lambda
https://github.com/developmentseed/geolambda
For instance, if you want to use the layer from lambgeo in us-east-1
, you can just add one of these arn
s to your function:
"layers": [
{
"name": "gdal24",
"arn": "arn:aws:lambda:us-east-1:524387336408:layer:gdal24:1",
"version": 1
},
{
"name": "gdal30",
"arn": "arn:aws:lambda:us-east-1:524387336408:layer:gdal30:1",
"version": 1
},
{
"name": "gdal31",
"arn": "arn:aws:lambda:us-east-1:524387336408:layer:gdal31:1",
"version": 1
}
For a full list of regions and arn
s, see this link.
After add layer, should add environment variables to lambda below informations
Here is an example using docker-lambda.
"GDAL_DATA": "/opt/share/gdal",
"PROJ_LIB": "/opt/share/proj",
"GDAL_LIBRARY_PATH": "/opt/lib/libgdal.so.2.x.x", // Enter correct version
"GEOS_LIBRARY_PATH": "/opt/lib/libgeos_c.so.1.xx.x" // Enter correct version
For django, edit your django settings.py
GDAL_LIBRARY_PATH = os.getenv('GDAL_LIBRARY_PATH')
GEOS_LIBRARY_PATH = os.getenv('GEOS_LIBRARY_PATH')