On Apr 27, 2009, at 2:19 PM, Brian Aker wrote:
Hi!
On Apr 27, 2009, at 12:07 PM, Sean Kelly wrote:
Feature Request: Get rid of table level write locks while hacking at
MEMORY engine.
Much easier to do if it lacks Btree from its current design :)
Yeah, but I could see BTREEs still being useful. My thought is that if
the table is really just a cache, memcached might be a better choice.
If a MEMORY table is being used, there's probably a (relational)
reason to use it, and, some of these cases, would involve BTREEs.
Doesn't PBXT do in-memory tables now? That sounds like a winner to me :)
Tim S.
_______________________________________________
Mailing list: https://launchpad.net/~drizzle-discuss
Post to : [email protected]
Unsubscribe : https://launchpad.net/~drizzle-discuss
More help : https://help.launchpad.net/ListHelp
