Closed bjtaylor1 closed 10 years ago
The output of ./osrm-extract on the server was:
[info] Input file: great-britain-latest.osm.pbf
[info] Profile: auk.lua
[info] Threads: 2
[info] Using script auk.lua
[STXXL-MSG] STXXL v1.3.1 (release)
[STXXL-MSG] 1 disks are allocated, total space: 1000 MiB
[info] Using turn restrictions
[info] Found 3 exceptions to turn restrictions:
[info] motorcar
[info] motor_vehicle
[info] vehicle
[info] Parsing in progress..
[info] Parsing finished after 605.008 seconds
[extractor] Sorting used nodes ... ok, after 5.446s
[extractor] Erasing duplicate nodes ... ok, after 1.155s
[extractor] Sorting all nodes ... ok, after 50.023s
[extractor] Sorting used ways ... ok, after 3.088s
[extractor] Sorting restrictions. by from... ok, after 0.143s
[extractor] Fixing restriction starts ... ok, after 0.431s
[extractor] Sorting restrictions. by to ... ok, after 0.147s
[extractor] Fixing restriction ends ... ok, after 0.056s
[info] usable restrictions: 5421
[extractor] Confirming/Writing used nodes ... ok, after 44.506s
[extractor] setting number of nodes ... ok
[extractor] Sorting edges by start ... ok, after 38.712s
[extractor] Setting start coords ... ok, after 69.102s
[extractor] Sorting edges by target ... ok, after 34.208s
[extractor] Setting target coords ... ok, after 85.628s
[extractor] setting number of edges ... ok
[extractor] writing street name index ... ok, after 0.056s
[info] Processed 12223401 nodes and 12768379 edges
[info] extraction finished after 938.011s
[info] To prepare the data for routing, run: ./osrm-prepare great-britain-latest.osrm
so I think it is an issue with prepare not reading them, rather than extract not creating them, as extract claims to have written them...
Looks like you ran out of disk space. Make sure /tmp has sufficient space. Alternatively use the latest code from develop branch which does not use /tmp anymore.
ok i'll give that a go. cheers
So I presume "develop" is the bleeding edge but less stable then? At what point do you tend to merge "develop" into main ? Any reasons why I might not want to use develop?
Develop is mostly stable. We are using it to power the demo site. As such there is quite some effort to make it as stable possible. It is merged into master branch once it has accumulated enough new features to justify a new release number.
For experiments and personal use develop branch is pretty alright. For production sites master is more tested.
ah. I seem to run into https://github.com/Project-OSRM/osrm-backend/issues/694 using develop. resolution seems to be to make sure profile.lua is a symlink to profiles/car.lua but still does it with that.
btw the build assembles the binaries into "Project-OSRM/build", I then copy them back one into "Project-OSRM" by doing "cp build/* ./", and run from there - is that the right way to do it?
I'd just symlink:
ln -s build/osrm-*
Hi, Just got the latest code and latest OSM data to update my routing servers. Seems to run ok on my own home machine, which is a laptop running ubuntu 14.
On my laptop osrm-prepare outputs:
But on my amazon server which is also ubuntu 14 osrm-prepare outputs:
(the last line is on stderr in red, and the exit code is 1). I presume the fact that it managed to find 0 edges as opposed to 12768379 is part of the problem?
Why could this be?
The great-britain-latest.osrm that they are using is not the same, on the server (which fails) it is 513,642,494 bytes in size but on my laptop (which is ok) it is slightly bigger at 565,857,571.
The only issue I did have updating the server was that it complained that it couldn't find something like "libboost_filesystem_1.46.so" so I got the latest boost 1.56, built it from source, installed it with "./bootstrap && ./b2 install", rebuilt OSRM, and then that error went away. I think my laptop is on boost 1.55 but still built from source.
Any ideas as to how I can track down this issue? The lua profile I am using is below in case it matters. But can't think it's an issue with that - otherwise it would surely not work on my laptop?