Re: CL Scaling for High Traffic Web Sites

On Apr 30, 10:08 am, Tim Bradshaw <tfb+goo...@xxxxxxxx> wrote:

The non-stupid design thing also matters. At some point I worked out
that we were averaging 1000 requests a second at the front end, well
over 90% of which were satisfied from cache, and the back end database
was sustaining 4000 IOPS. So that's worse than 400 IOPS per uncached
request. I kind of realised we were doomed at that point.

That's not the first time I've made that mistake: 40 IOs per uncached
request not 400, let alone 400 IOPS per request which not only is
wrong but doesn't make sense.

Sorry

.

Relevant Pages

  • Re: WriteFile()
    ... > In the case of two IOPs that write to overlapping regions of a file; ... > fact that you cannot know which request actually occurred first does not ... hardware device. ...
    (microsoft.public.win32.programmer.kernel)
  • Re: compact flash IDE hot-swap summary please
    ... but we shouldn't try to start a new request. ... Once we've flipped the iops it wont matter if we accidentally queue ... You need a call in it to zap the iops, 3 return codes so you can tell ... send the line "unsubscribe linux-kernel" in ...
    (Linux-Kernel)