How to run open-elevation to support 50,000 queries at a time

[kovar-ursa]

Putting this here as an issue as it should help others with the same question.

Change config.ini to:

port=8080 # Unless you want to run as root and can bind to 80
# How many GeoTIFF files to keep in memory for caching and request speeding up
open-interfaces-size=12
# Number of gunicorn workers
workers=8

Change server.py, right after "from gdal_interfaces"

import bottle
bottle.BaseRequest.MEMFILE_MAX = 1000 * bottle.BaseRequest.MEMFILE_MAX

Then run it with:

python server.py

It may be possible to tweak the config.ini better but those values work for me to support 50,000+ queries per call running locally.

If someone wants to build a new docker instance, that'd be great.

[vrzh]

Thanks for this. I was also getting the "Invalid JSON" error message and your fix of increasing bottle.BaseRequest.MEMFILE_MAX fixed it for me.

[aliasfoxkde]

Very cool but I think you might be bottlenecked by the API through web here. Alternatively, I took only the raw "gdal_interfaces.py" and stripped down the server.py file to get additional performance. But you can just import it directly into python directly "from server import *" and you can make about 50K queries per second (depending on hardware I imagine). I'm using a hybrid of the Web API and direct access myself.