Since two years or more I do from time to time a planet import into my postgresql databases on two servers because I needed additional tags by ways of the style sheet. Last successful one was in mid-December. When doing an import in January with just a few new tags, I recognized the import rate of nodes/second was only half of what I was used to. When I let it run through, my usually sufficient 2TB disk wasn't large enough. Replaced it with a 4TB disk but also this one ran out of disk space. Tried it with newest and some older osm2pgsql versions to no avail. As usual I think I haven't changed anything on the servers (Debian Buster)... Has anyone else every saw such a behaviour and figured out what was screwed up? Thanks. asked 17 Feb '21, 10:24 hkucharek |
Noticed in Osmose that the update backlog went from 0.4 days gradually towards the evening to 0.6 days yesterday. Just now it dropped to 0.4 days again, suspecting the US went to sleep. Rendering refreshes in ID editor carto and carto-fr is also very slow, some elements wont show for multiple days to a week. I guess that with everyone/many/most being cooked up in house due COVID-19, the use of OSM has gone up exponentially at times 'the site is under heavy load' popping up at times. In short, yes things seem under duress. answered 17 Feb '21, 11:10 SekeRob |