Hide Forgot
Description of problem: My system has 4 HDs, 3 SATA are in a RAID and used for Fedora (sda, sdb, sdc), the fourth is a PATA drive with a Windows install on it (sdd). normally sdd is not mounted under Fedora and, as it is old, I want it in standby mode. sda, sdb and sdc should not be in standby (those 3 drives are rated for 24x7 operation) When gnome-power-manager is running, sdd will be woken up from time to time. If I kill gnome-power-manager the disk seemingly remains in standby (it was this morning and every time I checked yesterday evening) Version-Release number of selected component (if applicable): gnome-power-manager-2.32.0-3.fc14 How reproducible: when gnome-power-manager is running the disk will invariably wake after a while but I am not sure it is that which is responsible for the wake Steps to Reproduce: 1. make sure smartd is started with '-n standby,10,q' 2. make sure hddtempd is not active for sdd 3. do not have 'spin down disks' active in gnome-power-manager as I want it to not touch sda, sdb and sdc 4. /sbin/hdparm -S 120 /dev/sdd 5. /sbin/hdparm -y /dev/sdd or /sbin/hdparm -Y /dev/sdd Actual results: sdd goes to standby, after a while it is active again. (checked with /sbin/hdparm -C /dev/sdd ) Expected results: gnome-power-manager, if it is the one waking the disk, should not wake a drive it has not put to sleep itself. Additional info: How can I be sure it is really gnome-power-manager that is waking the HD?
While this is similar to bug 656393 I do not think it is a dupe as that bug wants gnome-power-manager to put drives to sleep, whereas I want it to just leave my drives alone
I'm not sure what it's doing. If you had a laptop it's conceivable its writing battery stats data every few hours, but not on a desktop. You can try doing: killall gnome-power-manager gnome-power-manager --verbose then putting the disks to sleep, and then waiting for them to spin up. If g-p-m does anything major, it'll show up in the logs.
Done that and of course the waking did not happen again :-( I'll leave this NEEDINFO on me for now, but if within 1 month I do not manage to catch the machine out waking that disk this bug can be closed NOTABUG
still have not caught the box waking the disk. closing for now, will re-open if and when I notice further wakeups and manage to track down what wakes the disk