On Thu, Mar 24, 2011 at 04:05:07PM +0000, Eduardo Horvath wrote: > On Thu, 24 Mar 2011, Andrew Doran wrote: > > > I have a suspicion that SSD could cause issues for us because the buffer > > cache and other parts of the I/O system are not designed with near > > instantaneous request->response in mind, but that likely isn't at play > > here.. Do you have logging turned on for the SSD? > > Do you have a pretty standard SSD or something that's write optimized? > While it's probably not the cause of your performace problem since you're > spinning in the kernel rather than idle waiting for an interrupt, SSDs are > usually optimized for reads. While they have zero seek times, the write > cycles tend to be slow compared to spinning media.
As I said: only the sources are on SSD. And they have been preloaded into the page cache by tar cf - /usr/src > /dev/null before the build. The obj, destination, and release directories are on tmpfs. So I am hopeful this has nothing to do with the disk subsystem! -- Thor Lancelot Simon [email protected] And now he couldn't remember when this passion had flown, leaving him so foolish and bewildered and astray: can any man? William Styron
