Blast from the past time! Back in 2014 I filed Bug 14090 against Mageia 4 (and consequently, 5) complaining how my IDE hard disks were being left in a low power state after boot even with the box plugged into AC power. Well, after doing a clean install of Mageia 6, this behaviour has returned. I was pleased (though surprised, given how my research for that old report suggested it was being phased out) to find that the MGA6 repositories still contain the pm-utils package. Since pm-powersave is still available, I have -- for the time being -- reverted to the workaround I mentioned at https://bugs.mageia.org/show_bug.cgi?id=14090#c4 to mitigate the issue. That last time around, José Jorge squashed this bug properly with a udev rule ( https://bugs.mageia.org/show_bug.cgi?id=14090#c6 ). Is that still an option? I imagine pm-utils is still verging on deprecation and that using it is, by now, no longer the preferred or recommended route. So what to do? José, I hope you don't mind me adding you to the CC list for this report. Perhaps you could work some similar magic again? How reproducible: Always. Steps to Reproduce: 1. While it's hooked to AC power, boot your system. 2. Open a terminal as root. 3. # hdparm -B /dev/sd? Example of my output from step 3 (the values of which result in excessive spin-down of the drives): [root@purgatory ~]# hdparm -B /dev/sd? /dev/sda: APM_level = 128 /dev/sdb: APM_level = 96 Workaround: 1. Install the pm-utils-1.4.1-10.mga6 package. 2. In a terminal (even as a regular user) type in: $ pm-powersave 3. New results (no spin-down, fast access time): [root@purgatory ~]# hdparm -B /dev/sd? /dev/sda: APM_level = 254 /dev/sdb: APM_level = 254 For a more persistent workaround see https://bugs.mageia.org/show_bug.cgi?id=14090#c4
CC: (none) => johnltw, lists.jjorge
I don't know why pm-utils are not default installed anymore. But pm-powersave changes drive settings when a power event happens. So you should not need to trigger it manualy. When it it run, it logs to /var/log/ can you ensure that this file is not changed after a boot?
I don't know what/whom to assign this bug report to. CC'ing kernel and basesystem maintainers.
CC: (none) => basesystem, kernel, marja11
Assignee: bugsquad => lists.jjorge
CC: (none) => storage
On both my laptops (connected to AC power permanently) after installing Mageia 6 (XFCE) the hard disks started clicking like crazy because an APM level of 1 was set in both cases. Installing pm-utils afterwards didn't make any difference. I solved it in the old way by adding "hdparm -B 254 /dev/sda" to /etc/rc.d/rc.local.
(In reply to José Jorge from comment #1) > I don't know why pm-utils are not default installed anymore. > > But pm-powersave changes drive settings when a power event happens. So you > should not need to trigger it manualy. When it it run, it logs to /var/log/ > can you ensure that this file is not changed after a boot? Hi José. The only way to get the drives running at APM 254 is to kick pm-powersave into action via personal intervention (i.e. manually or by scripted workaround). I compared /var/log/pm-powersave.log immediately after boot and again after invoking pm-powersave, but there's really no change worth noting (the journal commit and readahead setting lines for my /home partition got moved around, but that's it).
The question is : does MGA6 change the APM of hard drives automaticaly? I don't think so. Maybe you can try to ensure this way : - booting MGA5 Live - booting MGA6 Live - booting any other distro And compare the APM state.
Ever confirmed: 1 => 0Status: NEW => UNCONFIRMEDKeywords: (none) => NEEDINFO
(In reply to José Jorge from comment #5) > The question is : does MGA6 change the APM of hard drives automaticaly? > > I don't think so. Maybe you can try to ensure this way : > - booting MGA5 Live > - booting MGA6 Live > - booting any other distro > > And compare the APM state. I'm not sure you understood the matter of this problem.
Status: UNCONFIRMED => NEWEver confirmed: 0 => 1
@José - I'm in agreement with Omnio (Comment #6) and baffled by your Comment #5, especially considering we've been through this all before with Bug 14090 -- which *you* fixed! The current bug appears (at least from a user perspective) completely identical to the former. Also, just so you know, since around the time of the recent KDE4 base upgrade (2017-08-24 - Bug 17123) your fix for Mageia 5 has been nullified as well. I'm guessing some update clobbered your udev rule. I've reinstated by old workaround on my remaining MGA5 machines and really don't want to reopen Bug 14090. :-(
(In reply to John ten Wolde from comment #7) > @José - I'm in agreement with Omnio (Comment #6) and baffled by your Comment > #5, especially considering we've been through this all before with Bug 14090 > -- which *you* fixed! Thank you for reminding me that. In fact it is my fault : I fixed in MGA5 forgetting to apply the fix in Cauldron, so now it misses in MGA6. Please test the uploaded pm-utils to ensure it fixes at least for Omnio. I don't know how the KDE4 upgrade could clobber this fix, at least it didn't in my system, I have just tested switching AC now. Suggested advisory: ======================== With systemd, the pm-powersave scripts were not launched anymore on laptops when ac-power starts and stops. This updated pm-utils brings an udev rule to start powersave on battery, on stop it on AC. This update done at first for MGA5 through bug 14090 was missing in MGA6. ======================== Updated packages in core/updates_testing: ======================== pm-utils-1.4.1-10.1.mga6.i586.rpm pm-utils-devel-1.4.1-10.1.mga6.i586.rpm Source RPMs: pm-utils-1.4.1-10.1.mga6.src.rpm
To test just watch the time of the file /var/log/pm-powersave.log on a laptop. It should change each time the ac power changes.
Keywords: NEEDINFO => (none)Status: NEW => ASSIGNED
Assignee: lists.jjorge => qa-bugs
MGA6-32 on Asus A6000VM MATE No installation issues. I can confirm the time on the file /var/log/pm-powersave.log changes every time I plugin or unplug the power line of the laptop. OK.
CC: (none) => herman.viaeneWhiteboard: (none) => MGA6-32-OK
(In reply to José Jorge from comment #8) Hi, José. These are the results of my tests: - When my laptops are plugged into AC power the APM_level is set to 254 as it should. This problem is solved. Thanks. - When running on batteries the APM_level is set to 1 and this is a way too aggressive power saving causing the hard disks to "click" (spin down / spin up - ad infinitum with only a few seconds interval). I had a look at the Ubuntu Mate install on one of my laptops and they set the APM_level to 128 when running on batteries, which is a more appropriate setting in my opinion.
(In reply to Omnio Torr from comment #11) > - When running on batteries the APM_level is set to 1 and this is a way too > aggressive power saving causing the hard disks to "click" (spin down / spin > up - ad infinitum with only a few seconds interval). I had a look at the > Ubuntu Mate install on one of my laptops and they set the APM_level to 128 > when running on batteries, which is a more appropriate setting in my opinion. Please open another bug report about that, to not hijack this one. Still, from the hdparm man page you can see that an APM value of 128 forbids the disk to spin down, so Ubuntu seems to just disable real APM on battery. And then spin down time is another parameter, just look the file /usr/lib/pm-utils/power.d/harddrive.
(In reply to José Jorge from comment #8) > Thank you for reminding me that. I had mentioned it in comment #1. ;-) > Please test the uploaded pm-utils to ensure it fixes Yes sir, I can confirm that did the trick. All good here on an x86_64 Mageia 6 laptop as well. Thank you.
Whiteboard: MGA6-32-OK => MGA6-32-OK; MGA6-64-OK
What do you guys think, should I open another bug and ask for pm-utils to be installed by default? It's good to have a working pm-utils package but for someone to benefit from it (s)he has to know about it and install it (and it seems the power management is really NOT OK without it at this very moment). I'm not a big fan of Ubuntu but just as an example I should mention they install pm-utils by default as a dependency of the meta-package ubuntu-mate-core.
(In reply to Omnio Torr from comment #14) > What do you guys think, should I open another bug...? Sure. If you'd like, go ahead. I concur both with your statement above, that pm-utils should be installed be default, and with what you said back in comment #11 about APM 1 being too aggressive. Perhaps an APM between 64 and 96 would be more sane, I really don't know, except that I agree with what José implied in comment #12 about Ubuntu "faking" power management. Spin-down (APM <128) *should* be enforced while on battery. As far as I can remember, Mandriva/Mageia[1] has always reduced APM to 1 on battery so I've grown to regard that as "normal", but whatever the setting, I suppose it's really a trade off between battery life, prolonged HDD health, and performance... If you do file another report, please do post its bug number here so I can add myself to its CC list. [1] I've never been into distro hopping. I started with Mandrake Linux 9.0 in 1999. Loaded Mandriva 2007.0 onto my first x86_64 laptop. A decade later I'm still using that same machine with Mageia 6. ;-)
Keywords: (none) => advisory, validated_updateCC: (none) => sysadmin-bugs
An update for this issue has been pushed to the Mageia Updates repository. https://advisories.mageia.org/MGAA-2017-0087.html
Resolution: (none) => FIXEDStatus: ASSIGNED => RESOLVED