Just came across the ddl of a table, which is as follows:
CREATE TABLE db.tab_1 ( item_id CHAr(10), Vendor_id INTEGER, .. ..., rptg_dt date format 'YYYY-MM-DD' ) PRIMARY INDEX xnupi_tab_1(item_id) Partition by Range_N(rptg_dt between date '2010-01-01' and date '2010-12-31' each interval '1' Day,Unknown);
Can anyone explain the partition key and the "Unknown" word in the Partition clause?
Hello, I have been looking for answers to a basic question on table partitioning in Teradata. Can someone please shed more light. Here it is -
If we do Range Partitioning on a table based on Numeric values and then different jobs load data corresponding to different partition ranges in parallel - then do we get into a resource contention scenario OR do all parallel running jobs hit specific partition and only acquire lock on a particular partition. In short - the question is - while loading data into a partitioned table [in this case Range Partition], do we acquire a lock at the partition level and multiple programs can run in parallel acquiring lock at individual partitions OR the lock is acquired at whole table level and the parallel jobs get into a contention mode.
Ex: An error table where each program loads specific error information that is encountered based on error code. If the table is range partitioned on error codes of range of 1000s, will individual programs when run in parallel hit their specific partition acquiring only lock to it or they will acquire complete table lock and release them for subsequent jobs when error load is done.