On Tue, 2009-11-10 at 15:42 +0000, Tyson Williams wrote:
> This proposed update did not speed up my system (that uses a SSD).
>
Ok, SSD gains are in the order of quarter to half a second - it's good
to know it hasn't regressed :p (this update is all about HDD
performance, not SSD)
> In all of these boot charts, you can see that there is about a 30
> seconds period where the CPU is barely being used, the CPU is waiting
> for the disk, the disk is being "utilized", but the throughput of the
> disk is basically zero. What is going on during this time? If it could
> be removed, I could at least get back to my ~30 Jaunty boot times.
>
hdparm sync of death. I should file a bug about that ;)
- delete /lib/udev/rules.d/85-hdparm.rules
Scott
--
Scott James Remnant
<email address hidden>
On Tue, 2009-11-10 at 15:42 +0000, Tyson Williams wrote:
> This proposed update did not speed up my system (that uses a SSD).
>
Ok, SSD gains are in the order of quarter to half a second - it's good
to know it hasn't regressed :p (this update is all about HDD
performance, not SSD)
> In all of these boot charts, you can see that there is about a 30 rules.d/ 85-hdparm. rules
> seconds period where the CPU is barely being used, the CPU is waiting
> for the disk, the disk is being "utilized", but the throughput of the
> disk is basically zero. What is going on during this time? If it could
> be removed, I could at least get back to my ~30 Jaunty boot times.
>
hdparm sync of death. I should file a bug about that ;)
- delete /lib/udev/
Scott
--
Scott James Remnant
<email address hidden>