I have a huge dataset and I want to split it into partitioned table, but I'm wondering, the second column is date in int and I wonder is it possible to split table based on this, because I don't want to create three more columns with year, month and day.
Yeah, i mean my table definition is something like this: fact1,fact2, date eg. (20190213), fact 4. What I want to do, is to create partitioned table based on year, month and day, but without creating three more fields with year, month and day. I worry that in approach of first creating "working" table and then putting it into final table it might get too slow, as everyday it's huge amount of data.
Yeah I think it should work, but correct me if I'm wrong, in the new table, date will be at the end of table? Also, I'm trying to partition it based on year then month then day, and I wonder if I can do it just by substring, and without dividing data field into 3 fields.
Yeah I saw that, but I'm not really sure if it will help me, as I want to first do partition by a yeah then month then day so the path would be like this /year/month/day/file so for example if person that runs query only know year and month it won't go through entire database.