This is a static archive of our old OpenStreetMap Help Site. Please post any new questions and answers at community.osm.org.

osm2pgsql stalled?

Hi,

I'm doing a full planet ingest using osm2pgsql on an Amazon AWS instance. I've given it a 25 gb cache using slim mode, and it's been very speedy. I started the process yesterday morning and now we're at the "Going over pending ways" portion.

I'm concerned however that the process has stalled out. My log file has not updated for 2 hours and I am stuck at

    Going over pending ways...
    191006558 ways are pending

    Using 1 helper-processes

And that's the end of my log file. When it was processing the nodes/ways/rels, at least I got an update of how far it was, but now it seems to have possibly stalled since nothing has updated in 2 hours.

How can I tell how far along this portion of the process is? Or, how can I tell if it has stalled out?

osm2pgsql

asked 11 Jun '15, 16:23

pixelk8e
26●1●1●2
accept rate: 0%

Did you already check CPU usage and I/O load of your osm2pgsql process?

(11 Jun '15, 18:02) scai ♦

Yes, it is using ~10-20% of the cpu at any given time (averaging at about 14%), but i/o seems to be nonexistent (I'm running osm2pgsql to a database on another instance, so that makes sense to me). But I'm still concerned it's chugging away at nothing.

(11 Jun '15, 18:17) pixelk8e

What does "wa" (I/O wait) say in top?

(13 Jun '15, 19:52) scai ♦

One Answer:

There are current versions of osm2pgsql that don't provide any output during that stage.

For estimated times for import see https://wiki.openstreetmap.org/wiki/Osm2pgsql/benchmarks

Note that due to OSM data growth these are always out of date (but can give you a rough indication).

answered 12 Jun '15, 18:58

SimonPoole ♦
44.7k●13●326●701
accept rate: 18%