[ale] offtopic: MySQL question

[Vernard Martin]

A colleague of mine is doing some research on backup technologies under linux.
He has run into a problem though:

>        In order to simulate an efficient restore algorithm (resore form
>backup tape), I am storing data about files that I back up in an msql
>database.  As far as I can tell, msql uses sockets to transmit commands to
>the msql2d process.  The problem is that to initalize the simulator, I
>must insert 85000 records.  However, because each insert waits for a
>reply, there is no pipelining, and I can only process 20 records per
>second.
>
>        One solution would be to use a different database.  One that
>would support sending batch of inserts, or one that I can link right into
>may application (so that it runs in the same address space).  
>
>        The other solution is to hack the msql source code and add another
>feature.  Unfortunately, it is OO code written in C, so there are many
>side-effects to most changes.

so the question is does anybody know how to get msql to do batchinserts or at
least speed up the inserts that it does? He doesn't want to switch from linux
if at all posible.

Vernard
--------------------------------------------------------------------------------
Darkness has a hunger that's insatiable   Vernard Martin (vernard at cc.gatech.edu)
Lightness has a call that's hard to hear  http://www.cc.gatech.edu/~vernard/