Importing data very slow after first import

General talk
1

Hi all!

I’m following this HowTo https://switch2osm.org/serving-tiles/manually-building-a-tile-server-debian-11/
Yesterday I started the import of Germany’s data. After 58 minutes done. Wow!
Then I imported the shape files.

Today I started the import of Czech. After more than one hour it didn’t start processing the nodes, ways and relation…
I find this is very strange, since Czech is very little than Germany…

Could someone explain me the behaviour?

The PC I’m using to process these data has a Xeon E5-2680 v2 2.80GHz (20 cores) and 32GB RAM. The disk I’m using for Postgres is an SSD.

Thanks for your explanation
Luca

2

Is this a new database or do you try to add the Czech data to the existing data for Germany?

3

I’m trying to import the data for Czech in the same DB I already used for Germany. So using --append instead of --create.
Data for Germany are in the DB.

Other question: I see now that the import for Germany didn’t complete without error:

node cache: stored: 330157901(100.00%), storage efficiency: 50.23% (dense blocks: 1515, sparse nodes: 322447601), hit rate: 100.00%
Osm2pgsql failed due to ERROR: Connection to database failed: FATAL: die verbleibenden Verbindungen sind für Superuser auf Nicht-Replikationsverbindungen reserviert

So someone know what does the program mean and how to solve it?!?

Thanks
Luca

4

You should have combined the files before the import e.g. with osmosis.

5

OK, can you say me how to do that (maybe an example or an HowTo)?

Then: what does the option “–append” mean, if not for this case?

Thanks
Luca

6

–append is used for updates over the original file (diffs). They come with osc extension.

About merging files, here’s a good starting point: https://help.openstreetmap.org/questions/5389/merging-2-countries-with-osm2pgsql

7

OK, thank you very much!

I’ll try to merge the data

Thanks
Luca