Re: Deletes in one schema slowing down reads in others

[Sean Maurik]

Thanks Jay but unfortunately I don't control the statement generation to that level (abstraction layer does it and is a bit out of my control). I'm pretty sure that efficient locking isn't my issue as its whole other schemas that are experienceing the slow down (though obviously if the "purge" statements is schema1 were to run faster themselves then the length of time that schema2 was slowed down would be improved but if the purges were constant than regardless of how fast they run I'd still be out of luck).

The key_buffer is interesting as thats the keyname in the sample .cnf files that shipped with the binaries (the my-*.cnf files sample configs) :). I'll try fiddling with the innodb buffer sizes and see if they help things at all.

Thanks
Sean