The Geek Stuff

Extracting data from a Notion database into your Oracle Autonomous Data Warehouse is easier than you think with the new RESTful API.

When it comes to hashing, what's fastest? RAW datatypes or VARCHAR2. I decided to conduct a few tests to find out. In this blog post I compare the two datatypes for insert and disk consumption.

A follow on from my previous blog entitled "The Hashed Speed Test & How to Avoid Collisions", I'm comparing DBMS_CRYPTO with Oracle's new 12c hashing function STANDARD_HASH on speed and reliability.

Oracle offers several hashing methods. In this blog post I review the four main types. I test the speed of inserts with hashing and which ones stand up to collisions (where two inputs hash to the same output).

I needed to insert 10s of millions of records into a table the other day for a future blog I'm putting together on hashing. For the life of me I couldn't work out how to achieve this, without using CONNECT BY ROWNUM (which produces errors running out of private memory). Finally cracked it, so thought I would share with the world.

In this quick demonstration I show the performance benefits of using bind variables in dynamic SQL.

I recently purchased “Expert one-on-one Oracle” by Tom Kyte. Packed full of years of knowledge from the export himself, one of the first sections I dived into was on tuning strategies and tools. In particular, SQL_TRACE, TIMED_STATISTICS, and TKPROF. As Tom alludes to, tuning tools are not used enough, which he believes is generally down to ignorance of their existence. I know I fall into that bucket.

A simple query rewrite example using materialised views and dimension metadata to control how the optimiser can tune SQL statements.