U
user
Can anyone in this group help me find out why Hibernation is so flaky?
P
Paul
http://support.microsoft.com/kb/899975/en-us
"Even though the system appears to be idle, the Windows XP System
Idle Task Scheduler service may be running system maintenance tasks.
These tasks may trigger a reset of the System Idle counter. In this
scenario, the computer does not enter standby.
Note Third-party programs and services may also be running during
system idle time. These programs and services may use more than
10 percent of the computer's CPU and disk resources. This level of
CPU and disk utilization may reset the System Idle counter and
therefore delay the system from entering standby."
Sounds pretty flaky to me.
I see two possibilities. Use keyboard/mouse activity, as a means
of determining a user is present. But that's not what people are
always looking for. Using %CPU isn't a particularly good
method, since it would not take much of a program to create a
spike that resets the timer.
As long as the system hibernates reliable when directly commanded
to do so, I'd say it is "mostly working". If you want to solve
this "hibernate under inactivity" problem, perhaps it'll take a
third party tool (or some policy adjustments), to get the
thresholding done with more precision.
I see claims that multimedia programs, have the ability to
"hold off" state changes, while things like movies are playing.
So that is yet another possibility. I can't imagine too many
multimedia players, managing to stay under 10% forever though.
Maybe they stay awake by just spiking occasionally
Paul