#DateForumTypeThreadPost
295325 Jul 2014 @ 01:22 PDTDatabaseReplyDate_Insert_errorIt's like any other switch in database technology: Doing it on your own might be a big task, especially if you don't have expertise in the new DBMS. There's 3rd party software like Sw...
295224 Jul 2014 @ 11:26 PDTDatabaseReplyDate_Insert_errorPlus 13.0.0.0 sounds like an ancient TD Express, you should install a newer version. Are you shure there's a TO_TIMESTAMP function? This was introduded in TD14, soy you must have an self-devel...
295124 Jul 2014 @ 09:58 PDTCloud ComputingReplyExpress 15.00.8_Sles11 worked fine for 2 weeks then "RDBMS CRASHED or SESSION RESET Recovery in progress"Something killed your AMPs :-( The reason should be somewhere in the messages, too. I don't know if this can be fixed or you need to use a new VM, you hopefully kept the ZIP file...
295024 Jul 2014 @ 09:54 PDTDatabaseReplyregexp_substr in TD14.10Current rleases of TPT support quoted data, you need to use following attributes for the DataConnector: SourceFormat = 'Delimited' TextDelimiter = ',' QuotedData = '...
294924 Jul 2014 @ 09:12 PDTDatabaseReplyerror in UPDATE statementThe syntax for UPDATE FROM is a bit strange: UPDATE t1 FROM DB_UTILS.T_TABLE_C_SIZE t1, ( select databasename, tablename, sum(currentperm) corrente from tablesize group by databasename, table...
294824 Jul 2014 @ 09:06 PDTCloud ComputingReplyExpress 15.00.8_Sles11 worked fine for 2 weeks then "RDBMS CRASHED or SESSION RESET Recovery in progress"Hi Robert, what's returned by pdestate -a? And check /var/log/messages for Teradata related messages
294724 Jul 2014 @ 08:25 PDTDatabaseReplyerror in UPDATE statementYou probably run that on an ancient Teradata release :-) Scalar Subqueries like this where not allowed in Updates before TD13 
294624 Jul 2014 @ 08:23 PDTDatabaseReplyNeed help on a case scenarioYou're close, you just need to add a condition, this should work: ,COALESCE(CASE WHEN B.DATED_ST >= C.DATED_ST THEN B.PRICE END, C.PRICE) It's much easier with the included DATED...
294524 Jul 2014 @ 07:50 PDTDatabaseReplyNeed help on a case scenarioHi Robert, please post unrelated questions as a new topic.  [DECIMAL(18,0), NOT NULL]  is not valid in Teradata and the definition of the PK_Competitor_Identifier constraint is missing ...
294424 Jul 2014 @ 06:58 PDTDatabaseReplyNeed help on a case scenarioMight be possible doing two joins to table1 or using ROW_NUMBER, but you need to add more details: You only show two rows per id or country, I assume there might be older rows, too. If yes, is the...
294324 Jul 2014 @ 06:47 PDTDatabaseReplyDate_Insert_errorThat's an error indicating a column or table name longer than the maximum of 30 characters. Based on your DDL and Insert this error should not be returned, it must be caused by something else....
294224 Jul 2014 @ 06:37 PDTDatabaseReplyProblem with Polish special characters in WHERE clauseHow did you insert those polish characters into a LATIN column? Most eastern european diacritical characters are not included in Teradata's LATIN.   Regarding your result, works fine fo...
294124 Jul 2014 @ 05:43 PDTDatabaseReplyregexp_substr in TD14.10This works for your example: select 'a,b,c," abc,def,dfg",m,n,"hi,how,are you",g,k' as x ,trim(both '"' from regexp_substr(x ,'("[^"]+&...
294023 Jul 2014 @ 08:03 PDTDatabaseReplyCreating a view from a CTE queryHi Richard, a CTE in a View is not allowed (don't ask me, why?). As it's not a recursive CTE you can re-write it using a Derived Table: create view hif_user.cprd_modexac5_atb_fnl as ( ...
293923 Jul 2014 @ 07:57 PDTDatabaseReplyCommon Table Expressions (CTE) Regarding the huge query int CTE #1 as source for the the recursive CTE #2: Check if the optimizer resuses the result of #1 in a spool instead of re-calculating it for every recursive step. If not...
293823 Jul 2014 @ 07:48 PDTDatabaseReplyCommon Table Expressions (CTE) Hi Dominic, according to the manuals you should be able to do that. But when I asked why a similar example didn't work, the answer was "it's known, currently the workaround is to def...
293723 Jul 2014 @ 02:09 PDTDatabaseReplyGroup by on a Group by not working as expected!!!Hi Peter,  of course it's a bug, all queries return the same (correct) result on a 15.00.00.03 :-) 
293622 Jul 2014 @ 11:41 PDTDatabaseReplyGroup by on a Group by not working as expected!!!Hi Peter, "query is not working" was a bit vague, now it's clear :-) If there's no error, but a wrong result set, it's a bug. You should contact your DBA immediately to subm...
293522 Jul 2014 @ 02:57 PDTDatabaseReplyHow to check what values are passed to procedure call from qrylog table ?DQBL has been enhanced in TD15 to write this info into a new dbc.DBQLParamTbl: BEGIN QUERY LOGGING WITH PARAMINFO
293422 Jul 2014 @ 01:10 PDTDatabaseReplyCorrect usage of Teradata12 joins is not that much, the same for 500k rows. It might be your query which leads to a bad plan. Could you show your query and the plan produced by Teradata (simply add EXPLAIN before the SELE...
293320 Jul 2014 @ 08:22 PDTGeneralReplyReplace a String in column Check the definition of your OReplace UDF, it might not work with Unicode columns or the maximum size for the input or output column is smaller then the size of the requesttext.  There will b...
293220 Jul 2014 @ 05:17 PDTGeneralReplyReplace a String in column Does your site follow the recommendation to create 1:1 views for every base table? Then there should be a process already to create those and you might simply add the access lock to it. When ther...
293120 Jul 2014 @ 04:51 PDTViewpointReplySytem in Degraded Mode in ViewpointLogged on to Viewpoint as an admin go to the Admin menu -> Systems -> Setup -> System Health - Thresholds  You can easily find that info in the Viewpoint User Guide, Chapter 3 System...
293020 Jul 2014 @ 04:46 PDTDatabaseReplyBlank space check based on the positionCan you show exactly what you tried? Of course this compares to spaces as expected. If not, the "blanks" might only look like spaces but are something different.  Can you check wit...
292920 Jul 2014 @ 04:36 PDTDatabaseReplyRandom sampling - ORDER BY HASHROW versus ORDER BY NULLThe ACCOUNT_ID is probaby the PK of you table, so you order by the hash of a a unique column, of course this returns the same result over and over again. But when you ORDER BY NULL (or 0 or 4...

Pages