Re: [Firebird-devel] Transactional Metadata (was: Planning the post v3 development)

[Dimitry S. <sd...@ib...>]

29.04.2014 12:48, Alex Peshkoff wrote:
> It's about reading table data from page cache when prepare gets slower
> many times. If system table is not in page cache it will be hundreds
> times slower.

   Yes, this particular piece of code. But how big % of execution time it takes in greater 
picture?

> If you think that I give too pessimistic estimation here please compare
> searching data in btree index + analyzing data page with record version
> check + analysis for GC (all this done locking appropriate pages for
> read) on one side with finding record in an array by index on another
> side (that's how metadata cache works).

   Of course finding of record in array by index is faster. But array is not versioned. 
Current metadata cache schema can be happily used on transaction level if transaction has 
snapshot IL. It is only RC which is going to be a problem.

> You will see that provided estimation is correct.

   Look at list_stayng(). I'm afraid that with this monster in background, all other 
expenses are ignorable.

-- 
   WBR, SD.