|
You are here |
sqlmaria.com | ||
| | | | |
www.bobbydurrettdba.com
|
|
| | | | | We have been discussing query tuning at work and I've come back to an example of the Oracle optimizer choosing the wrong plan due to an error in its calculation of the number of rows returned... | |
| | | | |
www.dbaglobe.com
|
|
| | | | | A blog about on new technologie. Hands-on note about Hadoop, Cloudera, Hortonworks, NoSQL, Cassandra, Neo4j, MongoDB, Oracle, SQL Server, Linux, etc. | |
| | | | |
connor-mcdonald.com
|
|
| | | | | Ever heard of a BAND join? It's not actually new, it has been around since 12.2. Jonathan Lewis wrote up a nice post on it was back in 2017 but in a nutshell MERGE join done without a band join SQL> select 2 e1.last_name|| ' has salary roughly between '|| e2.last_name comparison 3 from 4... | |
| | | | |
dban00b.wordpress.com
|
|
| | | 2016-11-29 EDIT: Oracle has announced that in 12.2 the features that drove this overuse of the result cache (dynamic stats triggered by SPDs) will default to off, and provided patches to apply to 12.1 as well. More info here. Also, some investigation by Mauro Pagano around the impact of the new OPTIMIZER_ADAPTIVE_STATISTICS parameter turned up... | ||