I have a table which has around 25-30 million rows and 15 columns. most of them are static or gets realtime updates. But for some of the columns(3 to 4) I get the data from hdfs once in a day and updates it which means these colums have to get updated once in a day .
I tried creating a new table from the hdfs output and doing join on the main table but that seems to be taking forever. I tried doing batch updates to it but that too is taking like too long.
What is the best way to do this ?
I am using postgres as my db