#DateForumTypeThreadPost
39408 Jan 2014 @ 08:34 PSTDatabaseReplyStatistics Creation on Tables and Stat Subsets to Enable Faster Join QueriesHi, The best way to check the required stats is to use help stats diagnostics. You can use it as follows: DIAGNOSTIC HELPSTATS ON FOR SESSION; Explian Your_Query Then go to the end of expl...
39308 Jan 2014 @ 08:19 PSTDatabaseReplyOREPLACE oops, I thought you were asking about the string length. Well, according to the documentation "The maximum length of a VARCHAR or CLOB result value is the maximum length that Teradata support...
39208 Jan 2014 @ 03:17 PSTDatabaseReplyQuery with TOP and COUNTRatnam,   Without order by clause the database will return any 10 rows from the table, but this result may not be truely random. The random function does extra processing to generate a rando...
39108 Jan 2014 @ 02:59 PSTToolsReplyBTET/ANSI Transaction in TPT/FASTLOADSan, Sure, you can do that test, but remember here the rows are inserted into error table due to data error, and this will fail transaction in case the errors exceed the defined error_limit. &nbs...
39008 Jan 2014 @ 01:29 PSTDatabaseReplyOREPLACE   JF,   According to the documentation the purpose of OREPLACE is to replace a part of string with another, to get the max string size you can use:   Select ...
38908 Jan 2014 @ 12:52 PSTDatabaseReplyFind out all the groups in a data.Barani, Your dataset is confusing, Do you want to find groups based on common To_Number and From_Number?  
38807 Jan 2014 @ 08:27 PSTDatabaseReplyConverting Number to CharacterSo you want to suppress any trailing 0z, This will work for you: SELECT CAST(11 AS DECIMAL(18,6)) / 1000.00 (FORMAT 'GZ(I)DZZZZZ')  
38707 Jan 2014 @ 07:59 PSTToolsReplyBTET/ANSI Transaction in TPT/FASTLOADSantanu, In Teradata utilities there are two kind of tables to handle errors, one is to handle UV violations and the other is used to store rows with errors, these errors are usually related to da...
38607 Jan 2014 @ 07:51 PSTDatabaseReplyConverting Number to CharacterStephen, Basically in order to avoid this round off you need to devide by the same percision as the value. for example in case of 18750 we need to case it to decimal to get result as decimal and ...
38507 Jan 2014 @ 12:39 PSTDatabaseReplyConverting Number to CharacterHi Stiphu,   Try this: SELECT CAST(CAST(2400 AS DECIMAL(18,1)) / 1000 AS VARCHAR(6))AS div  
38403 Jan 2014 @ 10:29 PSTTrainingReplyRegarding Certification Exams - Pass %Ashoke,  Thats great, if you are working on a client side then I think there should be a proper way to get all that material. You can discuss it with your manager, he will surely guide you ho...
38303 Jan 2014 @ 10:26 PSTGeneralReplyQuestionCheeli, No, You cant switch from TD-12 to TD-14 like that, you will have to start from TE0-141 :)  
38203 Jan 2014 @ 10:25 PSTGeneralReplyQuestionMonita, I suggest you to start with TE0-141, because TD-12 certifications will expire and you will have to bridge at any cost.  
38102 Jan 2014 @ 10:50 PSTTrainingReplyRegarding Certification Exams - Pass %Ashok, I have heard that is a good book, You will have to purchase that. But if you are not an employee of Teradata or have not worked on Teradata then whats the benifit of certification?  
38002 Jan 2014 @ 10:36 PSTDatabaseReplyJoin two tables using like statementDieter, My bad, I ignored that condition, thanks for the correction :)  
37902 Jan 2014 @ 09:14 PSTDatabaseReplyJoin two tables using like statementvandeberg, Agree with you, using a function, especially a string function will not only increase spool and full table scan, but also will cause the optimizer to use a product join strategy to reso...
37802 Jan 2014 @ 03:47 PSTDatabaseReplyStrange behavior while moving data from Integer columns to Varchar columns I dont know why I am unable to paste the code, but In fact Teradata is implicitly converting the integer to Varchar, it happens as follows:   Integer 1 is convertes as ----...
37702 Jan 2014 @ 03:01 PSTDatabaseReplyJoin two tables using like statementHi caka,   While looking at your data, 12345 in T1 has three 12345/1,12345 /2,12345/123, So in either left or inner you will get three values against a single value in table1, whic...
37602 Jan 2014 @ 02:44 PSTGeneralReplyQuestionHi, Yes, you can only jump from TE0-123 to TE0-143, and then attempt exam from TE0-144 to TE0-147. Other wise you will have to take start from TE0-141. http://www.teradata.com/tcpp/teradata-cert...
37502 Jan 2014 @ 02:29 PSTToolsReplyWhen are secondary indexes required?Bliane,   I didn't get what you want to conclude from this, but the discussion whther to use SI on a PPI or not, and when it is accessed by the optimizer.   You can see in all the...
37402 Jan 2014 @ 02:20 PSTDatabaseReplyStrange behavior while moving data from Integer columns to Varchar columnsHi,   It is not the strange behaviour rather it is default behaviour. An when an intger is converted to varchar, it requires 11 characters for representation. So when you try to insert the i...
37302 Jan 2014 @ 02:01 PSTDatabaseReplyOrder of partition columns in MLPPIHi San,   The base of decisions in defining MLPPI is to improve access, so you need to consider access and then decide about which col should be at the top and which at the lowest level. &n...
37201 Jan 2014 @ 11:18 PSTDatabaseReplyPerm Space and Spool space Hi, When you assign spool space to a user it defines the maximum spool limit that a user can use, but allocated only at run time according to the requirement. So deleting users does not mean that ...
37101 Jan 2014 @ 10:40 PSTToolsReplyWhen are secondary indexes required?Hi San, I must appreciate the effort you have put to simulate the scenario. Once again I would say, the final decision is of optmizer to use which index at which scenario. for example even if opt...
37001 Jan 2014 @ 12:14 PSTToolsReplyWhen are secondary indexes required?Hi VandeBergB, The case you are refering is the best scenario of partition and this is the most optimal case in PPI. Secondary indexes are mostly recomended when the partition by column is not par...

Pages