pbxt-discuss team mailing list archive

[Vladimir Kolesnikov <vladimir@xxxxxxxxxxxxx>]

Hi,

Here's what the original MySQL comments say:

/*
Reading keys in random order is as fast as reading keys in sort order
(Used in records.cc to decide if we should use a record cache and by
filesort to decide if we should sort key + data or key + pointer-to-row
*/
#define HA_FAST_KEY_READ (1 << 5)

I think it's formally not true for PBXT, if by "reading in sort order" 
they mean index scan.And here's the comment in PBXT :

/* Since PBXT caches read records itself, I believe
* this to be the case.
*/
HA_FAST_KEY_READ |

I guess this flag was initially intended for the case when the engine 
maintains a linear list of index leafs, so index scan is O(n). This is 
formally true for InnoDB but I think this flag might make sense for PBXT 
as well because the of the generally more effective index caching as 
mentioned in the comment.

14.05.2010 10:26, Paul McCullagh пишет:
> Hi Brian,
>
> Good question, I am not sure why that flag is set. Maybe Vlad has an 
> idea...
>
> What does the flag mean?
>
> On May 13, 2010, at 11:16 PM, Brian Aker wrote:
>
> > Hi!
> >
> > I was noticing that PBXT supports HTON_FAST_KEY_READ, is this really 
> > the case?
> >
> > Thanks,
> > -Brian
> > _______________________________________________
> > Mailing list: https://launchpad.net/~pbxt-discuss
> > Post to : pbxt-discuss@xxxxxxxxxxxxxxxxxxx
> > Unsubscribe : https://launchpad.net/~pbxt-discuss
> > More help : https://help.launchpad.net/ListHelp
>
>
>
> --
> Paul McCullagh
> PrimeBase Technologies
> www.primebase.org
> www.blobstreaming.org
> pbxt.blogspot.com