[Tile-serving] [openstreetmap/osm2pgsql] Possible performance improvement (#1532)

[doskabouter]

It was something like 64GB Commit size, Working set around 45GB and some 20GB Shareable
On windows server 2008.
Unfortunately the process has finished now so not able to check with procexp to see what's going on exactly.

Only thing that I was thinking about is a possible late free/delete after all the work is done instead of after the cache isn't needed anymore.
But if you think it's already freed in the code at the correct moment, then it's probably the runtime/memory manager holding on to it because it might be needed again. If that is true then I'd suspect it's too much effort forcing the memory back to the os

-- 
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/1532#issuecomment-876688613
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tile-serving/attachments/20210708/b1ac02c8/attachment.htm>