Hi, Lars, Thanks for suggestions.
Your execute parallel is very interesting. Are there explanations or a documentation or a guide available? Regards, David On Mon, 28 Mar 2022 at 11:09, Lars Aksel Opsahl <lars.ops...@nibio.no> wrote: > >From: postgis-users <postgis-users-boun...@lists.osgeo.org> on behalf of > Shaozhong SHI <shishaozh...@gmail.com>Sent: Monday, March 28, 2022 11:42 > AMTo: PostGIS Users Discussion <postgis-users@lists.osgeo.org>Subject: > Re: [postgis-users] Cut extent box2d into small regular boxes, and iterate > through each box of features > >Hi, Lars, > > > >Would the following alternative work? > > > >In PostGIS, one can group by geometry. Based on this information, can > grid be generated? so that grid with 0 features can be skipped. > > > > Hi > > In our cases this would not solve our problems, because grid cells with > zero content does not cause any big performance issue problems, but we do > not need them so it's nice to get reed of them. > > But I think it's easier to have a complete grids coverage and instead vary > cell size based on content. If you have a set of 10 mill polygons and and > you make a grid so you can process them in parallel what's important is > that each job has about the same needs related to memory and CPU, and this > can be done by setting for instance a max a value for number of features pr > cell. (You can here also count number points in each polygon but that takes > more time and you can not base the grid cell creation on indexes) > > Lars > _______________________________________________ > postgis-users mailing list > postgis-users@lists.osgeo.org > https://lists.osgeo.org/mailman/listinfo/postgis-users >
_______________________________________________ postgis-users mailing list postgis-users@lists.osgeo.org https://lists.osgeo.org/mailman/listinfo/postgis-users