#DateForumTypeThreadPost
110314 Sep 2011 @ 09:41 PDTUDAReplyHow to remove non-numeric characters from a string? (Oracle TRANSLATE Equivlent function?)Contact your Teradata support, they got Unicode versions of the Oracle UDFs. Dieter
110209 Sep 2011 @ 08:47 PDTDatabaseReplyHow to change CASE NOT SPECIFIC to CASE SPECIFIC for PI fieldsI don't understand what you're trying to achieve. An Insert/Select? Could you post the SQL which is "not working"? Dieter
110108 Sep 2011 @ 03:38 PDTAnalyticsReply10065 WSA E HOST UNREACH:Hi Partha, just a comment on your blog: TPA is not an interface to Teradata, it *is* the database software :-) /etc/init.d/tpa start" just triggers the startup process of Teradata. Dieter
110008 Sep 2011 @ 03:24 PDTDatabaseReplyTeradata Table Naming Hi Darcy, it's there, you just have to find it :-) Database Limits -> Table and View Limits: "Maximum database, user, base table, view, macro, index, trigger, stored procedure, UDF, UDM, UD...
109926 Aug 2011 @ 01:34 PDTTeradata ApplicationsReplyHow to install Teradata on UBUNTU#1: Install VMWare Player on Ubuntu #2: download TDExpress VMWare image Dieter
109826 Aug 2011 @ 01:33 PDTDatabaseReplycreating temporal tableThe temporal stuff is a new feature in TD13.10, but you're trying it on TD13.0 :-) Dieter
109726 Aug 2011 @ 01:30 PDTDatabaseReplyTeradata Table Naming Hi Darcy, there's an appendix "Teradata System Limits" in the Database Design manual listing various limitations. Dieter
109624 Aug 2011 @ 01:45 PDTDatabaseReplyFinding next change in valueNo need for a DENSE_RANK, you just need to look at the next 166 row after a 173 like: select ..., ts_col, ts_col - min(ts_col) over (partition by ... order by ts_col rows between 1 ...
109511 Aug 2011 @ 08:59 PDTToolsReplyImporting with BTEQUSING is only for the next *request*, looks like you try to use it in another request. USING IN_LAST_UPDATE_DATE (VARCHAR(19)) SELECT F.PARNT_ID PARNT_PR_ID FROM FDS_ETL_TARGET.QLT_INVSTGTN_F ...
109411 Aug 2011 @ 08:47 PDTGeneralReplyDELETING DUPLICATES I NEED SQL QUERY IN TERADATABefore the Teradata-ROWID was disabled (for various reasons) you could have written a similar query, of course comparing *all* columns. But now there's no DELETE query to achieve what you want. ...
109311 Aug 2011 @ 08:44 PDTDatabaseReplyDefining a different logtable in FASTLOADNo, the LOGTABLE command is only available for MLoad/TPump/FExp Dieter
109211 Aug 2011 @ 08:41 PDTDatabaseReplymultiple inserts in one statementHi Carlos, of course this is Standard SQL, it's a "row value constructor" :-) Dieter
109110 Aug 2011 @ 07:57 PDTAnalyticsReplySeem to be getting a 2646 spool space error when there is no spool space problemIs there a PROFILE for that user? Then you have to look at MaxProfileSpool instead of MaxSpool: COALESCE(MaxProfileSpool, MaxSpool) Dieter
109008 Aug 2011 @ 10:19 PDTDatabaseReplyChar column stores data with spaces but gives result even without spacesHi Gagan, all data in a CHAR column is padded to the maximum size with blanks. And according to Standard SQL the shorter string is padded with blanks when comparing two strings of different len...
108904 Aug 2011 @ 01:42 PDTDatabaseReplyData Cleansing using SQL"shifted down by one" means there *is* a kind of order. Then it would be easy: select col1, min(col2) over (order by whatever rows between 1 following and 1 following) col3,col4 from tab qu...
108826 Jul 2011 @ 01:31 PDTDatabaseReplyTimeStamp value for every record inserted using ins/selectHi Grao, within a request the value of current_timestamp never changes, even if it runs for hours, this is Standard SQL. If you want a unique timestamp(0) for each row you could write a Stored ...
108725 Jul 2011 @ 12:53 PDTDatabaseReplyTimeStamp value for every record inserted using ins/selectYou have to add a column list: insert into tbl_A (a) sel a from tbl_B; Dieter
108622 Jul 2011 @ 11:31 PDTAnalyticsReplyHelp Needed in writing SQLselect extract(year from logdate), extract(month from logdate), tablename, avg(space) from tab group by 1,2,3 Dieter
108522 Jul 2011 @ 03:42 PDTDatabaseReplySelect on Select - scalar subqueriesYou try to submit that query on a pre-TD13 system. Scalara Subqueries are fully supported for TD13, previously they were restricted to WHERE/HAVING. Dieter
108419 Jul 2011 @ 02:21 PDTAnalyticsReplyneed SQL help (The Output does not contain correct number of rows)Hi Sree, check the table definition if there's a FOREIGN KEY WITH NO CHECK OPTION REFERENCES tableA(id) The optimizer might do a join elimination if tableA.id is defined as NOT NULL (i.e. alway...
108315 Jul 2011 @ 08:36 PDTDatabaseReplyExecuting a multi-statement requestHi Andrew, i just tried it, any CALL in ANSI mode fails for a .NET connection, when you switch to odbc it succeds. Checking the query log i found its adding a COMMIT to the CALL as a multi-stat...
108215 Jul 2011 @ 03:37 PDTDatabaseReplyDROP OLDEST PARTITION AND ADD NEW PARTITIONOnly ranges at the "ends" of your partitioning schema can be added/dropped, but you try to drop a partition inbetween. What's the actual source code of your PARTITION BY? Why don't you simply d...
108115 Jul 2011 @ 03:30 PDTDatabaseReplyExecuting a multi-statement requestHi Andrew, a Stored Procedure can only be called in the same session mode it was created, ANSI or Teradata, as there are some differences in transaction handling. You must switch to Teradata mode...
108014 Jul 2011 @ 11:37 PDTDatabaseReplyDROP OLDEST PARTITION AND ADD NEW PARTITIONYpou query for the oldest partition with data in it, but not the oldest defined partition. This is always partition 1: DROP RANGE WHERE PARTITION =1 But caution, the previous "partition 2" w...
107914 Jul 2011 @ 01:53 PDTDatabaseReplyHelp for stats!These are some queries i use: 1. This is the calculation used by Teradata administrator, i use a slightly different one to look for skewed tables: /*** Skewed tables greater than x (1) GB, m...

Pages