#DateForumTypeThreadPost
4311 Feb 2016 @ 01:55 PSTDatabaseReplyWrong Statistic Estimations of Optimizer Teradata 14.10?Thanks Dieter. I tried it with BETWEEN and you are right: the estimation is 1 row. I agree that it's better to overestimate than to underestimate, especially as a wrongly chosen join ...
4210 Feb 2016 @ 07:09 PSTDatabaseTopicWrong Statistic Estimations of Optimizer Teradata 14.10?Hi,   I have a row partitioned table with the following statistics, and execute below SQL: COLLECT STATISTICS COLUMN (COL1) ON TheDatabase.TheTable; SELECT * FROM TheDatabase.TheTab...
4120 Oct 2015 @ 12:05 PDTDatabaseTopicSurrogate Key SQL - Bad Performance Hi,   I have below SQL which leads to a product join because of the OR join conditions.  It updates several surrogate keys at once in the STAGETABLE, but for each surrogate key...
4018 Oct 2015 @ 08:39 PDTDatabaseReplyProduct Join in Teradata 15As far as I remember this behaviour is the same like in previous releases and not related to Teradata 15. Actually, one of the "features" which can lead to errors whcih are very difficult...
3915 Oct 2015 @ 09:42 PDTDatabaseTopicRisks of CANCEL ROLLBACK of a tableHi, I know that by canceling a rollback we will be loose the table and it will have to be re-populated.  Apart from this effect, are their any other risks (for example related to the who...
3814 Oct 2015 @ 11:46 PDTDatabaseReplyRe write a queryCan you post the EXPLAIN output please?  Can you give as well the top result rows of: SELECT O_TRAN_ID,COUNT(*) FROM xyz.#VAU_DETAIL GROUP BY 1 ORDER BY 2 DESC ? Thanks Roland...
3711 Oct 2015 @ 10:28 PDTDatabaseReplySpool Space Error while running a select queryIn order to help you, i would like to see the SQL statement.  - How many rows are in each of the tables being joined? - As you mentioned skew: you should check if the join columns have skew...
3608 Oct 2015 @ 12:19 PDTDatabaseReplyIN Clause There is no difference.
3507 Oct 2015 @ 07:30 PDTDatabaseReplyALTER TABLE - Adding already existing columnsThe ALTER TABLE works, f you don't alter the column attributes and no statistics are defined on the column(s). 
3404 Oct 2015 @ 03:33 PDTDatabaseReplySliding Window Merge Join QuestionHi Kirthi, I tried out your ideas and finally, after increasing the table size from 70000 rows to millions of rows, the Optimizer finally switched to a row key based merge join in each single situ...
3302 Oct 2015 @ 11:12 PDTDatabaseReplySliding Window Merge Join QuestionWhat about the situation if the left table (NPPI) fits entirely into memory, and the right table is partitioned (PPI)? In this case each rowhash of the left table is compared against the first part...
3201 Oct 2015 @ 11:22 PDTDatabaseReplySliding Window Merge Join QuestionHi, I added the statistics you mentioned on both tables. Now i always get the traditional merge join with a redistribution step on (SalesId,SalesDate) for both tables at the begin...if i select al...
3101 Oct 2015 @ 08:06 PDTDatabaseReplySliding Window Merge Join QuestionSingle stats on SALESID (the PI) and SALESDATE (Partition column). We are on TD 14.10 Full Stats (no samples)
3030 Sep 2015 @ 11:55 PDTDatabaseReplySliding Window Merge Join QuestionI found another interesting behaviour when joining two tables with matching PI and Partitioning. I have the follwowing 2 tables:   CREATE MULTISET TABLE DWHPro.Sales2 ,NO FALLBACK , ...
2930 Sep 2015 @ 10:53 PDTDatabaseReplySliding Window Merge Join Question Thanks Dieter. Here are the table definitions:   CREATE MULTISET TABLE CDR.Big ,NO FALLBACK ,      NO BEFORE JOURNAL,      NO AFTER JOURNAL, ...
2829 Sep 2015 @ 10:43 PDTDatabaseTopicALTER TABLE - Adding already existing columnsI have a question: What will happen internally if i add columns with NULL values to a table which are already existing (exactly with the same attributes)? Is there any activity taking place, such a...
2729 Sep 2015 @ 10:35 PDTDatabaseTopicWrong resultset via JDBC in ANSI ModeI am not sure if the database forum is the right place for this topic, but it might be. I am selecting from a view which contains several OLAP functions. If the JDBC connection is set to Teradata ...
2629 Sep 2015 @ 10:21 PDTDatabaseReplyTactical WorkloadI checked for dictionary blocks, but the query was not blocked. But thanks for the hint, I will check if there was flow control during the execution of my query.
2529 Sep 2015 @ 07:39 PDTDatabaseReplySHOW Statistics - Meaning of AvgRowsPerBlock I wonder if this information is available somewhere else? I am asking because as far as I know is the usage of secondary indexes dependend from the average rows per data block(in order to determine...
2427 Sep 2015 @ 05:34 PDTDatabaseReplyPrimary index of PPI table has to be NUPI... Why is this so????The reason why UPI tables have to include the Partition columns as part of the Primary Index is that without this restriction, several PI values could be spread across different partitions. Wh...
2327 Sep 2015 @ 04:54 PDTDatabaseTopicSliding Window Merge Join QuestionHi, I have a huge table with billions of rows which is joined against a table containing 350.000 rows. Bith tables have the same PI and the same Partitioning. Statistics are available and correct....
2227 Sep 2015 @ 04:50 PDTGeneralReply1 MB data block size in 14.10I have a question which may fit this topic. Assuming we have very large table rows, let's say only a few rows fiting into one data block of 127KB. If i apply an ALTER TABLE statement, adding se...
2127 Sep 2015 @ 04:22 PDTDatabaseTopicTactical WorkloadHi, I have a question regarding tactical queries. I designed some tactical queries with index (USI) access to the rows. If i run this query on an unloaded system, response time is between 1-2 seco...
2027 Sep 2015 @ 04:13 PDTDatabaseTopicALTER TABLE and Workload Management Session ABORTHi, I encountered a strange situation regarding workload management. We are running a surrogate key generation stored procedure which does an ALTER TABLE to add the empty surrogate key columns at ...
1927 Sep 2015 @ 04:04 PDTDatabaseTopicSHOW Statistics - Meaning of AvgRowsPerBlock Hi, Could somebody help me with the AvgRowsPerBlock column returned in SHOW STATISTICS.  I have a table with  large columns (each about 64KB) and a datablock size of 256KB. On avera...

Pages