InstantForum.NET v4.1.4Teradata Forumshttp://www.teradata.com/teradataforum/info@teradata.comSat, 21 Nov 2009 19:02:53 GMT20http://www.teradata.com/teradataforum/Topic16006-9-1.aspx1.- Duplicate row check works different with INSERT...SELECT (remove duplicates silently) and INSERT INTO...VALUES (raise errors)2.- If there is a UNIQUE restriction on a SET table (UPI, USI...) Teradata uses this restriction (less costly) instead of the 'SET' duplicate row check.More information here (in spanish):http://carlosal.wordpress.com/2009/02/16/tablas-set-y-multiset-en-teradata/http://carlosal.wordpress.com/2009/07/06/tablas-set-y-multiset-en-teradata-ii/(Google translate is your friend...)HTH.Cheers.Carlos.Mon, 06 Jul 2009 05:04:31 GMTCarlosALhttp://www.teradata.com/teradataforum/Topic16006-9-1.aspxHI,Can anybody clear this please???Sat, 04 Jul 2009 23:31:06 GMTprakharhttp://www.teradata.com/teradataforum/Topic16006-9-1.aspxHi Carlos,I am bit confused as which one of you is saying correct ,also i was unable to understand anything as all in spanishCould you please write something from that extract in english explaining how really teradata checks SET command i.e mechanism involved in it as it is important for me to present all aspects in front of team to remove those 100 tables from SET to MULTISETRegards,Sat, 04 Jul 2009 05:15:37 GMTprakharhttp://www.teradata.com/teradataforum/Topic16006-9-1.aspxThis may be not true for INSERT...SELECT's. If there is an UPI on a SET TABLE, Teradata silently removes duplicates BEFORE the INSERT and there will be no 'Duplicate unique prime key error' (thus the 'duplicate checking' occurs no matter if there is an UPI or not).It is explained here (in spanish) http://carlosal.wordpress.com/2009/02/16/tablas-set-y-multiset-en-teradata/HTHCheers.Carlos.Fri, 03 Jul 2009 08:57:57 GMTCarlosALhttp://www.teradata.com/teradataforum/Topic16006-9-1.aspxIf a table has any unique index (primary or secondary) there is never any duplicate row checking. Checking for duplicate rows would be redundant in that case, so the Teradata software recognizes that it is not necessary.Thu, 02 Jul 2009 12:02:11 GMTJim Chapmanhttp://www.teradata.com/teradataforum/Topic16006-9-1.aspxI am stuck with a strange situation ....I am having about 100 tables in my system which areSET+UPI defined upon thatSET----system level check to ensure no duplicate rows are thereUPI----table level check to ensure a particular column has no repeated values thus identifying a row uniquelyMy doubt is that will SET+UPI combination will do duplicate row checking or only single time checking will be done?Also what is the process that really goesi mean to say 1st SET is checked then UPI check is done or it's vice versa...Please clarify as i need to change all 100 tables if SET+UPi is a problem...Also please clarify how really SET checking and UPI checking is done by Teradata while inserting dataRegards,Thu, 02 Jul 2009 11:11:00 GMTprakhar