480 | 05 Jan 2009 @ 09:03 PST | Database | Reply | Count on number of cursors opened | A session can have a max of 16 open result sets (Aka responses). it doesn't matter whether you are doing it in a procedure or in another proc called from that. The tally goes at session level not a... |
479 | 03 Jan 2009 @ 03:58 PST | Database | Reply | Non equi join condition | it's difficult to suggest something without knowing the actual joins / cardinailty /PIs and the explains.Most of the time when you have a non-equi join in the condition, the optimizer will chose to... |
478 | 03 Jan 2009 @ 03:44 PST | Database | Reply | Exporting Requesttest | You can generate the "show" statements itself at run time and use it to extract the view definitions ...i.e. something like this.** Generate all the show statements..export file = showstmts.sql;SEL... |
477 | 02 Jan 2009 @ 04:15 PST | Database | Reply | Non equi join condition | if your PI consists of only the first column, then you (might) still get the benefit of getting an AMP local join. |
476 | 02 Jan 2009 @ 04:09 PST | Database | Reply | Exporting Requesttest | Why don't you export the output of SHOW SELECT * FROM MYDB.MYVIEW;that way all the views/tables/JIs will be pulled out as well. |
475 | 30 Dec 2008 @ 11:02 PST | Tools | Reply | Find and replace | If your data can have '.' only at the end, then you might want to use a simple TRIM function..ieTRIM(TRAILING '.' FROM col1) |
474 | 30 Dec 2008 @ 10:59 PST | Tools | Reply | Control characters in output file | try this ....export file = mydata;SELECT ......export reset; |
473 | 30 Dec 2008 @ 10:56 PST | UDA | Reply | Unicode to Latin | I would say it kind of depends, and yes it's generally bad idea to keep to different character sets.For example, since the latin columns are going to be converted to unicode, this means the optimiz... |
472 | 30 Dec 2008 @ 02:24 PST | Tools | Reply | Mload Hangs in acquisition phase | Is multiload reading data from file or fifo ? what does the logs say ? |
471 | 28 Dec 2008 @ 01:42 PST | Database | Reply | PPI error | It needs to check only in one partition, the partition that corresponds to col3='2008-12-11'That's because it's perfectly valid to have another (409, 33,col3) as long as that col3 is not '2008-12-1... |
470 | 24 Dec 2008 @ 12:37 PST | Database | Reply | PPI error | There was a long discussion on this already.http://www.teradata.com/teradataforum/Topic6553-9-1.aspx |
469 | 12 Nov 2008 @ 09:37 PST | Analytics | Reply | Bad character in format or data | can you post the DDLs ? probably the data types are diff ? |
468 | 11 Nov 2008 @ 11:49 PST | Analytics | Reply | Dynamic Cursor For | Depending on the details of what you are trying to accomplish, you might be able to do a insert/select into an interim table or so and then run your queries against it. (will work if the columns yo... |
467 | 11 Nov 2008 @ 11:16 PST | Tools | Reply | error using import | You are having problem with date's format in the input data file. Try explicitly formating the input variable. |
466 | 09 Nov 2008 @ 05:22 PST | Database | Reply | taraData substring function | you shouldn't be having any trouble executing that SQL in 6.1Are you executing the entire contents in a single go ? it looks like there's something in your SQL assistant before the recursive statem... |
465 | 09 Nov 2008 @ 03:00 PST | Database | Reply | taraData substring function | It's the exact steps I executed.What version of TD are your running ? |
464 | 09 Nov 2008 @ 01:16 PST | Database | Reply | taraData substring function | You can try recursive queries , or a plain UDF which picks the last word in the text.the recursive solution is like this ...CREATE TABLE MYADD(ID INTEGER,ADDRESS VARCHAR(100))UNIQUE PRIMARY INDEX(I... |
463 | 09 Nov 2008 @ 01:02 PST | Tools | Reply | Handling timestamp column during fast export | It doesn't help in giving a solution if we don't know what you are doing with that time stamp column in the fast export script.Does this same SQL run in queryman ?Are you doing some conversion / ad... |
462 | 09 Nov 2008 @ 12:56 PST | Database | Reply | Reset identity column counter | A note of caution on doing soft RI on staging tables. This will mean that if you try to validate referential integrity on staging tables by executing an SQL or so, you might end up getting wrong r... |
461 | 09 Nov 2008 @ 12:43 PST | Database | Reply | IDLE SESSION BLOCKING!!!!!!!!!!!!!! | Could also happen if those "idle" session had explicit TD begin transaction (BT statements) and was not committed (ET). |
460 | 04 Nov 2008 @ 10:09 PST | UDA | Reply | ADDING HOURS IN TERSDATA SQL | you can add the ( interval '18' hour ) to your timestamp |
459 | 03 Nov 2008 @ 09:23 PST | Database | Reply | T7658 error | The error for scenario (1) is valid, you need to pass a host variable to bind the o/p value (in the case of SQL assistant, it was automatically taking care of it).I am not sure why (2) errors while... |
458 | 03 Nov 2008 @ 08:43 PST | Database | Reply | Collect statistics using variable | you can't refer to attribute or object names using a variable.try this instead.(echo ".logon $env/$user,$pwd;database $tdbase;"cat $scriptdir/tablelist | sed -e 's/^/COLLECT STATISTICS ON /'echo ".... |
457 | 03 Nov 2008 @ 08:32 PST | Analytics | Reply | Improving query performance | Are these 4 columns part of an index ? (Covered queries are generally faster since they are read directly from the index subtable).The other questions is how wide these 80 columns are, it can be to... |
456 | 27 Oct 2008 @ 10:38 PDT | Database | Reply | Adding a interval to datetimestamp from another column | (BEGIN_DATE (TIMESTAMP(0))) + ( ADD_HOURS(INTERVAL HOUR) ) |