Closed nilsnolde closed 2 years ago
When I started with the same arguments I had the same issue. Removing OVERPASS_SPACE=1000000000
resolved it.
Oh boy.. thanks a bunch! So it’s not safe to use that setting? I’ll try around a little and see what the limits are without explicitly setting it the max memory
I just started using the API again, I did that weeks ago already and I remember it working, local and with overpass-turbo.
However, now it doesn't seem to work. I pulled your 0.7.55.9 image and have the following docker-compose:
it's a tiny channel island, so it takes only 2 mins to spin up. the following curl
curl "http://localhost:31942/api/interpreter?data=node(5137693924)%3Bout%20geom%3B"
to just request this node ̣ results in HTML with the messageruntime error: open64: 0 Success /osm3s_v0.7.55_osm_base Dispatcher_Client::request_read_and_idx::timeout. The server is probably too busy to handle your request.
. looking into the dispatcher logs I see lines like these:I'm not familiar enough with overpass to interpret the very frequent
waited idle for 8 cycles.
lines and I have no idea why it's having read timeouts for such a trivial query. I checked the OSM file from geofabrik, it's legit, I even spun up a valhalla container to see if another app can read the .bz2 file, the graphs built.do you have any idea @wiktorn ? probably I'm overlooking smth really basic :sweat_smile: