#DateForumTypeThreadPost
50323 Apr 2009 @ 05:06 PDTDatabaseReplyRecursion in a stored procedureHi Jesse,sorry for the missinformation, i didn't know about that restriction in SPs.The view i mentioned needs to fully explode the hierarchy, thus you can filter for a given root. This example use...
50223 Apr 2009 @ 11:53 PDTUDAReplyBULK INSERT in TeradataHi Anderson,there's no built in equivalent for MS BULK INSERT, but you might use a BTEQ-Job (or a FastLoad/MultiLoad) to load data.What's the number of rows and the layout of the input file (CSV or...
50123 Apr 2009 @ 11:39 PDTToolsReplyWhere to get - TTU 12.0 for LinuxHi Akshay,you need access to Teradata @ Your Service to download the Linux Utilities.Or ask your friendly DBA.If you use the Teradata Express demo version, you're restricted to Windows.Dieter
50023 Apr 2009 @ 11:36 PDTToolsReplyMultiple flat files loading into singletable using floadRemove the END LOADING in your first script.Copy the script and change the filename to the second file.Copy the script and change the filename to the third file, this time include END LOADING.Run t...
49921 Apr 2009 @ 02:20 PDTDatabaseReplyRecursion in a stored procedureHi Jesse,* Teradata procs do not support the "With Recursive" clause of SelectsOf course you can use WITH RECURSIVE in SPs, but only TD12 allows a SP to return a result set.* Teradata macros do not...
49821 Apr 2009 @ 01:46 PDTDatabaseReplyHelp on a query for AggregationHi Manjeeth,you probably didn't specify the PI, thus it defaults to a NUPI on the first column Act_dt.And you didn't specify MULTISET, so it defaults to SET.Lots of rows with the same value + SET t...
49726 Mar 2009 @ 01:35 PDTDatabaseReplyIssue with TOP and Coalesce(Max)I don't know, why it's returning nothing, i just know it's stupid to mix TOP 1 and MAX, because both do the same.Maybe it's a bug, but why trying to find that out, if you just remove the TOP and it...
49626 Mar 2009 @ 01:25 PDTTrainingReplycse statementsIt's hard to understand what you are trying to achieve.It might be some UNIONs:select '02' from tab where column1 is not nullunion allselect '03' from tab where column2 is not nullunion allselect '...
49526 Mar 2009 @ 01:20 PDTDatabaseReplyDictionary CACHEThe parser also spoils the dictionary when it assumes that the number of rows changed significantly, usually for insert select and delete/update using non-index columns.Dieter
49426 Mar 2009 @ 01:16 PDTDatabaseReplyInstalled but could not use itHi Al,you probably skrewed up your installation, TDGSS is not installed correctly.I'd suggest to uninstall the Teradata demo completely (maybe even check the registry for remaining entries and remo...
49326 Mar 2009 @ 12:56 PDTToolsReplyFast Export issue (Kindly Help)A typical reason for that error is a left-over logtable from a MLoad/TPump job, which has different columns.But in your log file there's no name for the logtable. This script should fail with a syn...
49226 Mar 2009 @ 12:48 PDTToolsReplytimestampselect cast(ts_col as time)Dieter
49122 Mar 2009 @ 05:07 PDTDatabaseReplyInstalled but could not use itHi Al,that document is on the Express DVD, i don't have access to that right now, but you'll find usernames & passwords in it.And i think you'll find it in Start - Programs - Teradata demo, too.The...
49020 Mar 2009 @ 06:09 PDTAnalyticsReplyself joinTo get rid of the 7547 you have to aggregate the data in TEMP1 before updating.Dieter
48920 Mar 2009 @ 06:06 PDTToolsReplyLoading a timestamp column with FastloadHi Susan,did you try an explicit typecast in your INSERT?:CHN_TS (TIMESTAMP(6), FORMAT 'YYYYMMDDBHH:MI:SS.S(6)')Dieter
48820 Mar 2009 @ 06:02 PDTDatabaseReplyComputing week in TeradataHi Sam,the best way to write a date literal is DATE '2009-03-18', this will always work.So there's no need for CAST ('2009-03-18' as date)And of course in your query you might simply use calendar_d...
48720 Mar 2009 @ 05:49 PDTDatabaseReplyInstalled but could not use itBoot.ini is a hidden file used by windows during startup. You probably don't have to care about it, i suspect you just didn't start Teradata :-)What do you mean by "I configured it"? The Express ve...
48612 Mar 2009 @ 04:35 PDTAnalyticsReplyself join"different id1,id2,date combination for the same group"This means your grouping columns are not only (attribute1,attribute2,attribute3), but (id2,date), too.As you didn't provide DDL and INSERTs, t...
48512 Mar 2009 @ 03:59 PDTUDAReplyconvert varchar to timestampOr just add a FORMAT:select cast ('20090310181010' as TIMESTAMP(0) FORMAT 'yyyymmddhhmiss')Dieter
48410 Mar 2009 @ 02:09 PDTDatabaseReplyConstraintsHi Rick,if Foreign Keys are implemented, you might use the dbc.RI* views.But most environments don't implement it :-(Another way might be checking for columns with the same ColumnName.Best way is p...
48310 Mar 2009 @ 02:04 PDTDatabaseReplypartition By column contains "NULL"OLAP function follow the same rules as aggregates and UNION: NULLs are treated equal.select Col#1,Col#2,Col#3group by 1,2,3Dieter
48210 Mar 2009 @ 02:01 PDTUDAReplyImproper Column Reference in the search condition of a Joined Table.Hi Nick,don't mix old style and ANSI style joins.Add a JOIN to the Blank_customer.Dieter
48110 Mar 2009 @ 01:57 PDTAnalyticsReplyself joinSorry, but i can't figure out, what rows and why should be returned.Could you rephrase your example?Dieter
48006 Mar 2009 @ 10:46 PSTDatabaseReplyTeradata, turning off query logging(DBC.QRYLOG) for SQL'sQ1: no Q2: yesQuery Logging is controlled on a User/Account Level and can't be disabled for "individual" queries.Please contact your friendly DBA for further assistance :-)Dieter
47906 Mar 2009 @ 10:42 PSTConnectivityReplyDifference Versions of Client and "Server" Teradata.The client utilities are usually backwards compatible, so just use the TD12 versions.In case of ODBC you might use older versions, too.Btw, what's a Teradata v7 or v8.5?There was V2R6.0/V2R6.1/V2R6...

Pages