Segmentation fault (core dumped) about osm2pgsql HOT 2 CLOSED

Segmentation faults are almost always a sign of not having enough memory. Especially with this happening later in the process when you used a bigger machine. Figuring out how much memory something will use or is using is not straightforward. It might well be that everything looks fine and the next thing the program needs twice as much memory as before. So I`d recommend testing with a larger machine.

You can also try with --disable-parallel-indexing --number-processes=1. This will make osm2pgsql do less stuff at the same time possibly reducing its memory footprint somewhat. (But processing will take longer.)

That being said, of course there could be a different problem there. If you want to pursue this, create a debug build and look at the core dump created by the segmentation fault, specifically we need to backtrace do diagnose further.

from osm2pgsql.

Thanks for the feedback @joto.

I've tested with a different machine with 16 Gb of memory and I didn't get a segmentation fault. It's hard to know if this really was a memory issue or not, since the machines could have been entirely different, but as you say, segmentation faults are almost always a sign of not having enough memory.

It's weird though, that running on a machine (bare metal) that only had less than 4 Gb of memory available didn't raise any issues. Maybe because it used less memory from the start..

Anyway, closing this issue.

from osm2pgsql.