Paul, I kindly received the information about the table data (quoting here): >It changes as it goes down the table, it's a right mixture. ST_LineString | 2 | 5398548 ST_LineString | 3 | 2877681 ST_LineString | 4 | 2160809 ST_LineString | 5 | 1696900 ST_LineString | 6 | 1362231 ST_LineString | 7 | 1107733 ST_LineString | 8 | 915616 ST_LineString | 9 | 766904 ST_LineString | 10 | 646150 ST_LineString | 11 | 550356 ST_LineString | 12 | 473357 ST_LineString | 13 | 410038 ST_LineString | 14 | 358185 ST_LineString | 15 | 313985 ST_LineString | 16 | 278846 ST_LineString | 17 | 248253 ST_LineString | 18 | 220736 ST_LineString | 19 | 198809 ST_LineString | 20 | 179552 ST_LineString | 21 | 162140 ST_LineString | 22 | 147957 ST_LineString | 23 | 134321 ST_LineString | 24 | 123805 ST_LineString | 25 | 113805 ST_LineString | 26 | 105329 ST_LineString | 27 | 96809 ST_LineString | 28 | 90105 ST_LineString | 29 | 83137 ST_LineString | 30 | 77846 ST_LineString | 31 | 72963 ST_LineString | 32 | 67830 ST_LineString | 33 | 63849 ST_LineString | 34 | 60241 ST_LineString | 35 | 56312 ST_LineString | 36 | 52805 ST_LineString | 37 | 49919 ST_LineString | 38 | 47402 ST_LineString | 39 | 44860 ST_LineString | 40 | 41987 ST_LineString | 41 | 40055 ST_LineString | 42 | 38173 ST_LineString | 43 | 36649 ST_LineString | 44 | 34464 ST_LineString | 45 | 32637 ST_LineString | 46 | 31695 ST_LineString | 47 | 29851 ST_LineString | 48 | 28546 ST_LineString | 49 | 27419 ST_LineString | 50 | 26784 .... ST_LineString | 1993 | 2 ST_LineString | 1995 | 1 ST_LineString | 1997 | 6 ST_LineString | 1998 | 5 ST_LineString | 1999 | 3 ST_LineString | 2000 | 9 ST_Point | | 20648939 ST_MultiPolygon | | 6188 ST_Polygon | | 8054680 >One thing you could try is indexing each geometry type in turn and >watch the memory usage. If indexing starts sequentially from the beginning to the end, we reach ST_Point around row 22.000.000. This is from the count of geometry operations exactly where the slowdown begins. Does this track us to the leak? Thanks & regards Frans 2010/3/26 Paul Ramsey <pramsey@xxxxxxxxxxxxxxxxx>: > > Could you > - put in your version information > - tell me what kind of spatial objects you have? polygons of > 100 > vertices? lines of two vertices? etc. That will help me pick similar > data for the memory testing. > -- Sent via pgsql-general mailing list (pgsql-general@xxxxxxxxxxxxxx) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-general
