[Tile-serving] [openstreetmap/osm2pgsql] Very slow node parsing on updates (caused by PG's jit) (#1045)

[Sarah Hoffmann]

The estimates for the way-node index have always been widely off. It The index is so huge and badly balanced that postgres cannot make sense of it. It just never mattered because the query is simple enough that postgres always figured that it should use the index as opposed to a sequential scan, which were the only choices then. We are running now into trouble only because postgres tries to be even more clever and use such advanced features as workers and jit.

There might be a way to influence the statistics. Newer versions of postgres have more advanced tools for that, see [Postgresql Extended Statistics](https://www.postgresql.org/docs/10/planner-stats.html#PLANNER-STATS-EXTENDED).

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/openstreetmap/osm2pgsql/issues/1045#issuecomment-570559087
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tile-serving/attachments/20200103/9b81d154/attachment.html>