hard disk power management after resume

[KyleWhittington]

I'm running Gutsy on my Dell Inspiron 6400 notebook. It was a dist-upgrade from Feisty. I have found an issue with the hard disk after I resume from standby. The hard disk seems to keep going on standby within a couple seconds of being idle, like literally 5-10 seconds. The result is that every time that I then type or open up a new webpage, I hear the hard disk spinning up again. I have tried setting the power management profile using hdparm, even setting it to disabled, but the hard disk is still going to standby and then spinning up again after a few seconds of inactivity.

I'm guessing that it's kernel related, though it could be settings. As I said, hdparm has no impact on it, and the readout for the current power management setting in hdparm is as follows:

        Standby timer values: spec'd by Standard, no device specific minimum
        R/W multiple sector transfer: Max = 16 Current = 8
        Advanced power management level: unknown setting (0x0080)

These values are all the same even after it has come back from standby. Setting the APM level to 255 (disabled) has no effect on changing this value either and, as stated above, does not correct the problem.

If I reboot the notebook the problem goes away, and I can use it without any hassles from the hard disk spinning up and down the whole time. As soon as I go to standby and resume, the problem occurs, and it's reproduceable every single time. This was never a problem with any release before Gutsy, and I can't seem to find any other bug-reports for it.

[Ulrik Mikaelsson (rawler)]

This bug has gotten some attention. http://www.linux-hero.com/rant/explanation-ubuntu-hard-drive-wear-and-tear

Is it, like suggested in the comments, /etc/acpi/power.sh that controls this? It reads /etc/default/acpi-support and in my config it says;
# Switch to laptop-mode on battery power - off by default as it causes odd
# hangs on some machines
ENABLE_LAPTOP_MODE=false

So, is this related to the acpi-support package, or is it more related to gnome-power-manager?

If acpi-support, making a patch that adds an option to the defaults, "LAPTOP_MODE_HARDDRIVE_POWER=X" with an explanation of the issue should be trivial, and a great bug for mentoring someone new. If a problem with gnome-power-manager, perhaps it's a bit more tricky.

[Square Bottle (squarebottle)]

For being such a high attention bug that affects everybody, why is this not getting any attention? Especially when they only current fix is to outright disable APM?

Especially in light of how many people were concerned by this (it was on Digg after all), this should be treated as a very important issue and somebody should be assigned ASAP. It would definitely not be good to have this reach even more news sites unsolved, especially since it was already brought up. I can already hear the comparisons to Windows' slow response times if that happens.