A comment I always make in my seminar about Oracle technology is that the best way to find bugs is to combine two different bits of the technology. It’s easy to say, of course, but how can you possibly guess which combinations are going to produce bugs.
Here’s a really surprising one: Bug no: 3469992 – fixed in 10.2 (Mar 2006) so probably no longer relevant to most people.
Rediscovery Information: If flashback query fails with ORA-600 and db_block_size is bigger than 8K, it is due to this problem. Workaround: using smaller block_size
How on earth are you supposed to guess that flashback query and blocksizes would have anything to do with each other ? After all, flashback query is just read-consistency from a different perspective – it couldn’t possibly go wrong !