PG normally splits table data into 1GB chunks. The number before the dot is called the filenode. You can translate it into a table name by
select oid::regclass::text from pg_class where relfilenode='2474695';
I believe there is an option to change that chunk size but you'd have to recompile Postgres. The setting you need to change is called segment_size.
On Fri, Nov 15, 2024 at 3:27 PM Andy Hartman <hartman60home@xxxxxxxxx> wrote:
I created a new table (V16) and then used SimplySql to take data from mssql to the new Postgres table. The table is 212gig in size. Myquestion comes from the files created on the OS(Windows2022 server) I can see lots of files with the last being:
2474695.143They are all 1,048,576kb
Is this normal behaviour and could I have done something to use fewer files and larger ones?
This table is created in a separate tablespace on a dedicated drive on the windows file system.
I'm just getting involved in this PostgreSql instance
THanks.
