Hi all,   Let me first describe my problem. I have a set of tables (let's say there are 40 tables), each table has more than 2 millions different customers. I have set of columns that have to be updated in every table. For example, in every table I have to update PESEL column in the following way: '111_customer_id' I have a list of customers to be updated (a lot of customers) I created my program in that way, that for every customer I am creating update rule and I run this rule. So for every table I run something like this: update customer set pesel='111_&ID.' where ID=101; update customer set pesel='111_&ID.' where ID=102; /* and so on */ But this solution takes a lot of time - and it is simply unacceptable.   Then I tried to change my program to use hash tables instead of simple update. But I also have problems: because I have to make separate rule for every client (there is client's ID in PESEL rule), I have to make temporary table and transpose it for every customer. Also, I cannot use hash tables for numeric columns (like birth_date), because after transposing every colum is character (so I have an error that types does not match).   So my question is: is there any 'tricky' solution to figure this problem out? Maybe I should use hash tables, but in different way? Or make some parallelism? Do you have any ideas?   I will be grateful for any help!   Cheers! ... View more