[Bug 545024] New: Rapidly increasing Load_Cycle_Count
http://bugzilla.novell.com/show_bug.cgi?id=545024 Summary: Rapidly increasing Load_Cycle_Count Classification: openSUSE Product: openSUSE 11.2 Version: Milestone 8 Platform: Other OS/Version: Other Status: NEW Severity: Critical Priority: P5 - None Component: Kernel AssignedTo: kernel-maintainers@forge.provo.novell.com ReportedBy: rhafer@novell.com QAContact: qa@suse.de Found By: Development Created an attachment (id=321512) --> (http://bugzilla.novell.com/attachment.cgi?id=321512) hdparm -I output On a 11.2M8 test system ("Shuttle X27D" equipped with an Intel Board and a ATOM 330) smartctl shows me a rapidly increasing Load_Cycle_Count of the Harddrive. It unloads almost once per second. The harddrive is only about a week old and already has a Load_Cycle_Count of above 40000. Disabling the harddrive powermangement (hdparm -B 255 or hdparm -B 254) has no effect on this it seems. Only keeping the disk busy prevents it from unloading. Find the output of hdparm -I and smartctl -a attached. Please tell me what more information you need. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 User rhafer@novell.com added comment http://bugzilla.novell.com/show_bug.cgi?id=545024#c1 --- Comment #1 from Ralf Haferkamp <rhafer@novell.com> 2009-10-07 09:39:44 MDT --- Created an attachment (id=321513) --> (http://bugzilla.novell.com/attachment.cgi?id=321513) smartclt -a output -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 User info@tristanhoffmann.de added comment http://bugzilla.novell.com/show_bug.cgi?id=545024#c2 Tristan Hoffmann <info@tristanhoffmann.de> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |info@tristanhoffmann.de --- Comment #2 from Tristan Hoffmann <info@tristanhoffmann.de> 2009-10-30 14:36:57 MDT --- I also noticed this issue after upgrading from 11.1 to 11.2 RC1 and I think this problem should be solved prior to the release of 11.2, because this could damage many discs. In my case the disc loads/unloads about every 5 seconds (WD Scorpio 500GB), which is still too much while on AC power. I did not have any similar problems with 11.1, so it should be possible to fix this, maybe by executing hdparm -B 255 while on AC. Thanks. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 Tristan Hoffmann <info@tristanhoffmann.de> changed: What |Removed |Added ---------------------------------------------------------------------------- Priority|P5 - None |P1 - Urgent Severity|Critical |Blocker -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 User coolo@novell.com added comment http://bugzilla.novell.com/show_bug.cgi?id=545024#c3 Stephan Kulow <coolo@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Priority|P1 - Urgent |P5 - None Severity|Blocker |Major Flag| |SHIP_STOPPER- --- Comment #3 from Stephan Kulow <coolo@novell.com> 2009-11-01 13:22:02 MST --- an online update will do - for now it's not even clear what's the problem. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024#c4 Alberto Passalacqua <albert.passalacqua@gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |albert.passalacqua@gmail.co | |m --- Comment #4 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2009-11-18 17:16:38 UTC --- This bug also affects my Dell XPS m1330, with an increase of the load cycles of about 1000 in 4 days, which corresponds to the number of days I have been using 11.2 I attach hdparm and smartctl outputs. Could we increase severity, since it is not nice to have this kind of problems that might damage our hardware? Best -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024#c5 --- Comment #5 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2009-11-18 17:17:34 UTC --- Created an attachment (id=328261) --> (http://bugzilla.novell.com/attachment.cgi?id=328261) hdparm -I -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024#c6 --- Comment #6 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2009-11-18 17:18:02 UTC --- Created an attachment (id=328262) --> (http://bugzilla.novell.com/attachment.cgi?id=328262) smartctl -A -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024#c7 Alberto Passalacqua <albert.passalacqua@gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Version|Milestone 8 |Final --- Comment #7 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2009-11-18 17:18:28 UTC --- Moving to final... -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c8 --- Comment #8 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2009-11-19 04:41:56 UTC --- I'm not sure if this can help, but the most significant increases in load cycles happens here when the system is completely idle, even when plugged. One example: I stopped working at my laptop at 8pm I went back two hours later, and the load cycles went from 4022 to 4128 in this period, during which the system was doing absolutely nothing. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c9 Ralf Haferkamp <rhafer@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Severity|Major |Critical --- Comment #9 from Ralf Haferkamp <rhafer@novell.com> 2009-11-19 07:49:41 UTC --- Depeding on the harddrive it seems that it is possible to workaround this problem with the hdparm. I just exchanged my "SAMSUNG HM500JI" with a "Seagate Momentus 5400.6" and after calling "hdparm -B 254" the Load_Cycle_Count doesn't increase any more. But as written in comment#0 this didn't help for the Samsung drive. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c10 Jeff Mahoney <jeffm@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |jeffm@novell.com AssignedTo|kernel-maintainers@forge.pr |teheo@novell.com |ovo.novell.com | --- Comment #10 from Jeff Mahoney <jeffm@novell.com> 2009-11-19 22:27:30 UTC --- Tejun, is this a kernel issue or a userspace setting issue? -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c11 --- Comment #11 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2009-11-20 00:14:09 UTC --- In reply to comment #9 hdparm -B 254 significantly reduces the load cycles increase on my laptop (I had 8 in one day). Btw, 254 is also the recommended value by the vendor of my harddrive for acoustic reasons, while the value picked by default by openSUSE is 128. hdparm -B 255 completely prevents the increase of load cycles on the same machine. How can I make the change permanent? Thanks in advance, Alberto -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c12 Tejun Heo <teheo@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO Info Provider| |rhafer@novell.com --- Comment #12 from Tejun Heo <teheo@novell.com> 2009-11-21 00:31:41 UTC --- This is a userland setting issue but I wonder who's setting the APM value to 128 to begin with? If the firmware is doing that, we need to add the entry to storage-fixup. If something else is doing it, we need to find out which and remove it as it can be destroy certain drives very quickly. Can you please cold boot into single usermode (specify 1 at the boot loader prompt) and attach hdparm -I output from there? Thanks. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c13 Alberto Passalacqua <albert.passalacqua@gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |NEW Info Provider|rhafer@novell.com | --- Comment #13 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2009-11-21 08:43:34 UTC --- Created an attachment (id=328816) --> (http://bugzilla.novell.com/attachment.cgi?id=328816) hdparm -I output at runlevel 1 Hi, I attach what you requested. According to the output, the value is still 128. Best, Alberto -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c14 --- Comment #14 from Tejun Heo <teheo@novell.com> 2009-11-21 11:07:05 UTC --- Well, that must be the default value then. Are you sure your harddrive wasn't doing the same thing w/ opensuse 11.1? Can you boot 11.1 live cd and check whether the APM value is 128 too? -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c15 --- Comment #15 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2009-11-21 16:36:33 UTC --- I will check, but I am pretty sure. This hard drive has about two years of life (8874 power on hours), and 4134 cycles. It had about 3500 immediately after I installed 11.2. I have been using openSUSE 10.3, 11.0, 11.1 on it. Best, Alberto -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c16 --- Comment #16 from Ralf Haferkamp <rhafer@novell.com> 2009-11-23 10:13:58 UTC --- Created an attachment (id=328927) --> (http://bugzilla.novell.com/attachment.cgi?id=328927) hdparm -I output at runlevel 1 (SAMSUNG HM500JI) Mine has: "Advanced power management level: disabled" after booting into runlevel 1. But as I mentioned already in comment#0. Setting hdparm -B 255 or 254 as no effect on the Load_Cycle_Count behaviour of that drive. I can't say how 11.1 behave wrt. Load_Cycle_Count the system is pretty new and I never had 11.1 running on it. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c17 Tejun Heo <teheo@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO Info Provider| |rhafer@novell.com --- Comment #17 from Tejun Heo <teheo@novell.com> 2009-11-23 17:11:21 UTC --- Ralf, can you please run "blktrace /dev/sda" for a minute and attach the output files here? -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c18 Ralf Haferkamp <rhafer@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |NEW Info Provider|rhafer@novell.com | --- Comment #18 from Ralf Haferkamp <rhafer@novell.com> 2009-11-24 08:30:43 UTC --- # blktrace -d /dev/sda -w 60 === sda === CPU 0: 128 events, 7 KiB data CPU 1: 340 events, 16 KiB data CPU 2: 52 events, 3 KiB data CPU 3: 0 events, 0 KiB data Total: 520 events (dropped 0), 25 KiB data Load_Cycle_Count increased by 16 during that run. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c19 --- Comment #19 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2009-11-30 21:56:25 UTC --- Hi, sorry for the delay in the reply. I could try with 11.1, I did *not* notice any significant increase in the load cycle in about 12 hours of use, while on 11.2, without issuing hdparm -B 255, the increase is quite rapid (and noisy!). Could you tell me how to reset the powermanagement permanently at boot? Thanks, -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c20 Tejun Heo <teheo@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO Info Provider| |rhafer@novell.com --- Comment #20 from Tejun Heo <teheo@novell.com> 2009-12-01 00:53:02 UTC --- Ralf, can you please attach the output file saved by blktrace? They're named DEVNAME.blktrace.CPU. Alberto, can you please do the same? Run blktrace for a minute while the system is otherwise idle and attach the resulting files? It would be best if you run blktrace while cd'd to a usb stick so that blktrace itself doesn't cause IOs. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c21 --- Comment #21 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2009-12-01 01:34:00 UTC --- Created an attachment (id=330164) --> (http://bugzilla.novell.com/attachment.cgi?id=330164) sda.blktrace.0 Here they are :-) -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c22 --- Comment #22 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2009-12-01 01:34:46 UTC --- Created an attachment (id=330165) --> (http://bugzilla.novell.com/attachment.cgi?id=330165) sda.blktrace.1 -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c23 --- Comment #23 from Tejun Heo <teheo@novell.com> 2009-12-01 05:48:37 UTC --- Does "echo 1 > /proc/sys/vm/laptop_mode" make the problem go away? -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c24 --- Comment #24 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2009-12-01 06:10:10 UTC --- Hi, I tried resetting hdparm -B to the value it has at boot (128) and then issuing the command you suggested. It did not work: 5 cycles in about 15 minutes :-( Thanks, Alberto -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c25 Tejun Heo <teheo@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |jack@novell.com --- Comment #25 from Tejun Heo <teheo@novell.com> 2009-12-01 06:14:33 UTC --- Jan, on Alberto's machine, after upgrading to 11.2, pdflush is constantly issuing writes which interacts badly with the drive head unloading. There isn't much which can be done from under the block layer. How do we find out who's causing those frequent writeouts? Thanks. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c26 Ralf Haferkamp <rhafer@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |NEW Info Provider|rhafer@novell.com | --- Comment #26 from Ralf Haferkamp <rhafer@novell.com> 2009-12-01 07:55:16 UTC --- Created an attachment (id=330195) --> (http://bugzilla.novell.com/attachment.cgi?id=330195) blktrace output files blktrace files from the blktrace run in comment#18. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c27 --- Comment #27 from Ralf Haferkamp <rhafer@novell.com> 2009-12-01 07:59:29 UTC --- (In reply to comment #23)
Does "echo 1 > /proc/sys/vm/laptop_mode" make the problem go away? IIRC I tried that already and it didn't have any significant effect. I can't test currently, as I put a different harddisk into the machine for now (one where hdparm -B 254 cures the problem).
-- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c28 Ralf Haferkamp <rhafer@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO Info Provider| |jack@novell.com --- Comment #28 from Ralf Haferkamp <rhafer@novell.com> 2009-12-01 08:01:03 UTC --- Setting needinfo to jack@novell.com to answer comment#25. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c29 Mark Gray <markgray+to-suse@puck.nac.net> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |markgray+to-suse@puck.nac.n | |et --- Comment #29 from Mark Gray <markgray+to-suse@puck.nac.net> 2009-12-01 10:41:39 UTC --- (In reply to comment #25)
How do we find out who's causing those frequent writeouts?
With 11.0 I found that it was udevd on my laptop for what it is worth (I have an old Armada with a loud hard disk right by my bed which kept me awake with the clicking.) Just my $0.02. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c30 --- Comment #30 from Jan Kara <jack@novell.com> 2009-12-01 12:17:10 UTC --- Sadly, it's quite tricky to find out which application does the writes. So far the best we can do is to find out which files the writes happened to and then guess which process might be guilty. We can try it in this case and see whether we can see something. To do this, you should capture the blocktrace when the machine should be idle on a separate harddrive (or USB stick or to tmpfs in memory). For example for filesystem on /dev/sda1 do: mkdir /tmp/fs mount -t tmpfs none /tmp/fs blktrace -d /dev/sda1 -D /tmp/fs <wait some time and break blktrace> Now process the blktrace output by: blkparse -i sda1 -D /tmp/fs -f "%a %S+%n %C\n" | grep "^D" | cut -d ' ' -f 2,3
/tmp/blktrace-parsed
Download my mapping program (I suppose you have ext2/3/4 filesystem) from http://beta.suse.com/private/jack/defrag/ext3remapper.tar.gz Untar it and run make (you will need a development version of libext2fs which is packaged in openSUSE, otherwise it has no dependencies). Run: e2block2map -t -o /tmp/blktrace-mapped /tmp/blktrace-parsed /dev/sda1 It will crunch for some time (a few minutes) and then write in /tmp/blktrace-mapped which files were read / written. You can then attach the file here. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c31 Jan Kara <jack@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Info Provider|jack@novell.com |rhafer@novell.com --- Comment #31 from Jan Kara <jack@novell.com> 2009-12-01 12:18:33 UTC --- Setting needinfo to reporter for data file. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
Download my mapping program (I suppose you have ext2/3/4 filesystem) from http://beta.suse.com/private/jack/defrag/ext3remapper.tar.gz
Untar it and run make (you will need a development version of libext2fs which is packaged in openSUSE, otherwise it has no dependencies). Run: e2block2map -t -o /tmp/blktrace-mapped /tmp/blktrace-parsed /dev/sda1 There is no e2block2map built all I got was e2block2file, e2remapblocks and
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c32 Ralf Haferkamp <rhafer@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |NEW Info Provider|rhafer@novell.com | --- Comment #32 from Ralf Haferkamp <rhafer@novell.com> 2009-12-01 14:24:31 UTC --- (In reply to comment #30) [..] preloadblocks. I guess you meant e2block2file? The output of that is: 4603+1 unmapped [pdflush] 4916+1 unmapped [pdflush] 5023+1 unmapped [swapper] 9629+1 0 INO 8118 /tmp [swapper] 1309+1 unmapped [swapper] 0+1 0 GRP 0: superblock or its copy [swapper] 1+1 0 GRP 0: group descriptor [swapper] 777+1 unmapped [swapper] 794+1 unmapped [swapper] 188261+1 101 INO 59779 /var/log/mail.err [kjournald2] 203188+1 180 INO 59776 /var/log/mail [kjournald2] 203699+1 179 INO 59778 [swapper] 209588+1 180 INO 59777 /var/log/mail.info [swapper] 1610890+4 4488 INO 8 [swapper] 1610894+1 4492 INO 8 [swapper] 4603+1 unmapped [pdflush] 4537+2 unmapped [swapper] -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c33 --- Comment #33 from Jan Kara <jack@novell.com> 2009-12-01 14:44:29 UTC --- Yes, I meant e2block2file. Sorry for the mistake. OK, sadly quite a bit of the IO are short-lived files (probably in /tmp) that get deleted before e2block2file is run. Anyway, the files we can identify are mail logs. I suppose you are running postfix. Is your disk still woken up so often when you stop postfix? Also you can try stopping cron because I remember that used to wakeup disk quite often as well. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c34 --- Comment #34 from Ralf Haferkamp <rhafer@novell.com> 2009-12-01 14:51:46 UTC --- (In reply to comment #33)
Yes, I meant e2block2file. Sorry for the mistake.
OK, sadly quite a bit of the IO are short-lived files (probably in /tmp) that get deleted before e2block2file is run. Anyway, the files we can identify are mail logs. I suppose you are running postfix. Yes, I noticed that as well. There was a config problem that caused postfix to log very 30 seconds. I'll stopped postfix for now and try again.
Also you can try stopping cron because I remember that used to wakeup disk quite often as well. Ok.
-- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c35 --- Comment #35 from Ralf Haferkamp <rhafer@novell.com> 2009-12-01 15:07:49 UTC --- Ok, I stopped cron and postfix and kept blktrace running for 2 minutes. Load_Cycle_Count only increased by 4 during this run. The e2block2file output is just this: 631569+1 unmapped [pdflush] -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c36 --- Comment #36 from Jan Kara <jack@novell.com> 2009-12-01 15:22:46 UTC --- Ok, thanks for debugging. We might find something more if you do dd if=<your device> of=deleted-file bs=4096 count=1 skip=631569 deleted-file should then contain the data of the deleted file and we might be able to guess what file it is. Anyway, it's a bit strange that Load_Cycle_Count increased by 4 when we see just 1 write. Are there other partitions on the device? If not, could you paste here output of blkparse on the captured trace so that we see whether my filtering didn't miss something important? -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c37 --- Comment #37 from Ralf Haferkamp <rhafer@novell.com> 2009-12-01 15:49:13 UTC --- (In reply to comment #36)
Ok, thanks for debugging. We might find something more if you do dd if=<your device> of=deleted-file bs=4096 count=1 skip=631569
deleted-file should then contain the data of the deleted file and we might be able to guess what file it is. Hm, it seems that it was just the e2block2file output file. I deleted that between two blktrace runs. Why does it show up in blktrace, when I deleted it when blktrace was not running.
Anyway, it's a bit strange that Load_Cycle_Count increased by 4 when we see just 1 write. Are there other partitions on the device? There are more partions yes. But none of them is mounted. And the swap partition is disabled currently.
If not, could you paste here output of blkparse on the captured trace so that we see whether my filtering didn't miss something important? Hm, I deleted it already. Will try to get a new trace ...
-- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c38 --- Comment #38 from Mark Gray <markgray+to-suse@puck.nac.net> 2009-12-01 16:09:06 UTC --- You might save time by also stopping udevd and see if the problem stops (you only need it after booting if you are hotplugging): quad: ... /udev-146 # find -name '*.c' -print0|xargs -0 grep unlink|wc 25 87 1422 The ones in 'udev-146/udev/udev-watch.c' are the most suspect to my mind. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c39 --- Comment #39 from Ralf Haferkamp <rhafer@novell.com> 2009-12-01 16:24:23 UTC --- Created an attachment (id=330296) --> (http://bugzilla.novell.com/attachment.cgi?id=330296) blkparse output Attached is the output of blkparse -i sda2 -D /tmp/fs after a 2 minutes blktrace run. The e2block2file output was empty. Load_Cycle_Count increased by 2. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c40 --- Comment #40 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2009-12-01 16:44:59 UTC --- Hi, not sure if this is right, since I used /dev/sda6 (/ partition) since sda1 and sda2 are Windows partitions, sda3 is a Linux extended partition containining /sda6 and others. Anyway: blktrace-mapped rm 518144+1 0 INO 6905 /var/tmp/kdecache-alberto/kpc/plasma_theme_openSUSEdefault.index [kjournald2] 2658222+6 2989 INO 8 [kjournald2] 2658228+1 2995 INO 8 [swapper] 2658229+1 2996 INO 8 [swapper] 352884+1 0 INO 10730 /tmp/kde-alberto/khtmlcachetj9091.tmp [kblockd/0] 2658230+5 2997 INO 8 [kjournald2] 2658235+1 3002 INO 8 [kwin] 2658236+1 3003 INO 8 [kjournald2] 2658237+6 3004 INO 8 [swapper] 2658243+1 3010 INO 8 [swapper] --- blktrace-parsed 0+0 kjournald2 124114073+8 kwin 0+0 swapper 124114081+8 kjournald2 124114089+48 swapper 0+0 swapper 0+0 swapper 0+0 swapper 124114137+8 swapper 0+0 swapper -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c41 --- Comment #41 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2009-12-01 17:08:17 UTC --- It's not my case, since I use KDE and pulseaudio should not be running, but someone on IRC reported this too: http://www.phoronix.com/forums/showthread.php?t=20377 I checked with atop and dmesg, and the only thing I could see was a huge number of disk access from kjournald and pdflush on my machine. Best, Alberto -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c42 --- Comment #42 from Jan Kara <jack@novell.com> 2009-12-02 15:41:01 UTC --- Ralf, as you've probably guessed, those last two wakeups are from some daemon probing the disk once in a while. I guess it's DeviceKit dk-disks daemon from the name. You can try stopping it and see whether we can get rid of the last wakeups. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
You can try stopping it and see whether we can get rid of the last wakeups. Yeah, I wonder what it is worth though. The result is that I have a machine
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c43 --- Comment #43 from Ralf Haferkamp <rhafer@novell.com> 2009-12-02 15:54:14 UTC --- (In reply to comment #42) that is mostly idle and has no services running :-). But as soon as I start doing something the Load_Cycle Problem is back. The machine was supposed to be a small server (running postfix, cyrus, ldap and a few other things). Am I right assuming that this Samsung harddisk is just not suited for such a setup as I can't even get it to stop loading/unloading using hdparm -B 254 (or 255)? Or can this somehow be worked around in software? -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c44 --- Comment #44 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2009-12-07 02:13:38 UTC --- I'm sorry if I ask again. How do I make the hdparm settings permanent? It is a bit annoying to be forced to have to type the command at every boot, and a solution does not seem close. Thanks, Alberto -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c45 --- Comment #45 from Tejun Heo <teheo@novell.com> 2009-12-07 02:18:23 UTC --- You can put it in /etc/init.d/boot.local. The setting may be lost over suspend/resume cycle. For that, you'll need to add configuration under /etc/pm/sleep.d/. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c46 --- Comment #46 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2009-12-07 02:23:12 UTC --- Hi Tejun, thanks! :-) -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c47 --- Comment #47 from Jan Kara <jack@novell.com> 2009-12-08 09:35:17 UTC --- Ralf, as far as I know there's no way to tell kernel to avoid doing IO to the drive when an application does it. So if you cannot tell the drive to stop unloading, then yes, it's not a good drive for your server. Tejun, or is there some other way that might work to make the drive stop unloading? -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c48 --- Comment #48 from Tejun Heo <teheo@novell.com> 2009-12-09 04:31:37 UTC --- I'm afraid not. Maybe we can implement something which monitors activity and issue a read command if the device stays idle for a while so that it never goes to sleep but I don't think it's something we can push upstream. A temporary solution would be running "dd if=/dev/sda of=/dev/null bs=512 count=1 iflag=direct" or whatever command to keep the drive loaded from userland but device idleness can't be detected easily from userland. :-( Thanks. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c49 Daniele Tombolini <kailed@kailed.net> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |kailed@kailed.net --- Comment #49 from Daniele Tombolini <kailed@kailed.net> 2009-12-13 10:56:15 UTC --- Same issue here on my acer 2930z. When in idle load_cycle_count increase every 5 seconds! With hdparm -B 255 instead of (254 default) things seems a lot better after a quick test but hdparm -I still reports 254. Racommended value is 128. I'll attach output of smartctl -a and hdparm -I soon. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c50 --- Comment #50 from Daniele Tombolini <kailed@kailed.net> 2009-12-13 10:58:51 UTC --- Created an attachment (id=332346) --> (http://bugzilla.novell.com/attachment.cgi?id=332346) smartctl -a output -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c51 --- Comment #51 from Daniele Tombolini <kailed@kailed.net> 2009-12-13 10:59:49 UTC --- Created an attachment (id=332347) --> (http://bugzilla.novell.com/attachment.cgi?id=332347) hdparm -I output before setting -B 255 -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c52 --- Comment #52 from Daniele Tombolini <kailed@kailed.net> 2009-12-13 12:26:31 UTC --- #49 Sorry, I did a bit of confusion in apm/acoustic management. Default APM level 128. With 254 it seems better but I have to keep an eye on HD temperature. I 'll try to lowering the apm value, for a lot of people 200 is enough. Old thread with a list of affected machines/drives: http://ubuntuforums.org/showthread.php?t=795327 -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c53 --- Comment #53 from Tejun Heo <teheo@novell.com> 2009-12-15 05:27:37 UTC --- Daniele, can you please post an email to linux-ide@vger.kernel.org and attach the above information along with measurement of how many load cycles happen per minute under idle condition? Oh... and please cc me. Thanks. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c54 --- Comment #54 from Daniele Tombolini <kailed@kailed.net> 2009-12-15 19:00:13 UTC --- #53 done ! Ah.. 12-13 cycles for minute, sorry forgotten in email.. For the record, hdparm -B 200 fix the issue. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c55 --- Comment #55 from Kris Kaiser <Kris.Kaiser@gmx.net> 2010-01-07 00:44:14 UTC --- With Samsung HM160HC: 4 Load_Cycle_Count per Minute. Solution: APM: Advanced power management level: 254 (255=Disabled did *not* solve the Problem) hdparm -B 254 /dev/sda => No more Increase in Load_Cycle_Count -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c56 Tejun Heo <teheo@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO Info Provider| |Kris.Kaiser@gmx.net --- Comment #56 from Tejun Heo <teheo@novell.com> 2010-01-11 08:04:28 UTC --- Kris, did that HM160HC come preinstalled or was it an upgrade? Thanks. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c57 Kris Kaiser <Kris.Kaiser@gmx.net> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |Kris.Kaiser@gmx.net --- Comment #57 from Kris Kaiser <Kris.Kaiser@gmx.net> 2010-05-17 17:02:38 UTC --- (In reply to comment #56)
Kris, did that HM160HC come preinstalled or was it an upgrade?
Thanks.
an Upgrade -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c58 Alberto Passalacqua <albert.passalacqua@gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |NEW InfoProvider|Kris.Kaiser@gmx.net | --- Comment #58 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2010-06-28 01:30:40 UTC --- The problem is present also on a brand new Latitude E6410 with openSUSE 11.3 RC1, with the following Seagate hard-drive: Model Family: Seagate Momentus 7200.4 series Device Model: ST9500420AS Serial Number: 5VJ68LZJ Firmware Version: D005SDM1 User Capacity: 500,107,862,016 bytes Device is: In smartctl database [for details use: -P show] ATA Version is: 8 ATA Standard is: ATA-8-ACS revision 4 In this machine the recommended value for the power management settings for acoustic management is, according to hdparm -I Recommended acoustic management value: 208, current value: 208 however openSUSE ignores it and set it to 128 (exactly as on my previous machine). The load cycle count increases much quicker than there with the default value: 1180 cycles in two days, before I checked. Setting hdparm -B 254 /dev/sda seems to fix the problem. I'll tell you what happens in a while, after more certain information. However, it would be nice that the appropriate settings would be chosen out of the box, since this problem has been around for a long while :-) FYI, the hdparm -I and the output of smartctl -a are attached. Additionally, I move this to openSUSE 11.3 RC1, since it affects it too. Best, Alberto -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c59 --- Comment #59 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2010-06-28 01:31:48 UTC --- Created an attachment (id=371987) --> (http://bugzilla.novell.com/attachment.cgi?id=371987) hdparm -I on Latitude E6410 with Seagate HD -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c60 --- Comment #60 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2010-06-28 01:32:36 UTC --- Created an attachment (id=371988) --> (http://bugzilla.novell.com/attachment.cgi?id=371988) smartctl -a on Latitude E6410 with Seagate HD -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c Alberto Passalacqua <albert.passalacqua@gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Priority|P5 - None |P2 - High Component|Kernel |Kernel Version|Final |RC 1 AssignedTo|teheo@novell.com |kernel-maintainers@forge.pr | |ovo.novell.com Product|openSUSE 11.2 |openSUSE 11.3 Target Milestone|--- |RC 1 -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c61 Tejun Heo <teheo@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |ASSIGNED CC| |teheo@novell.com --- Comment #61 from Tejun Heo <teheo@novell.com> 2010-06-28 05:46:17 UTC --- openSUSE doesn't diddle with APM setting by default. That's either the drive's default or BIOS configuring it during POST. The only thing that we can do is adding the configuration to storage-fixup as temporary workaround but storage-fixup matches individual configurations and it doesn't make sense to add an entry for upgraded hard drive there. We're currently talking with hard disk vendors to resolve the issue. Hopefully, the problem can be resolved from hard drive side. For now, I'm afraid you'll just have to add appropriate hdparm command to /etc/init.d/boot.local. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c62 --- Comment #62 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2010-06-28 07:13:59 UTC --- OK, I did that already. However this hard drive is the original one coming with the laptop, and not an upgraded hard drive. Additionally, some post ago you said the setting in boot.local could be lost if we use standby, and you suggested to add something in /etc/pm/sleep.d/. I do not have anything to edit in that directory. Could you please give me an example of what I should put there? Thanks in advance :-) -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c63 --- Comment #63 from Daniele Tombolini <kailed@kailed.net> 2010-06-28 20:21:36 UTC --- @Andrea, a simple bash script it's enough. Well, a bit quick & dirty and if you don't like, this is a full script (a template) for this kind of things: #!/bin/bash case $1 in hibernate) echo "Hey guy, we are going to suspend to disk!" ;; suspend) echo "Hey guy, we are going to suspend to RAM ;; thaw) echo "oh, suspend to disk is over, we are resuming..." ;; resume) echo "hey, the suspend to RAM seems to be over..." ;; *) echo "somebody is calling me totally wrong." ;; esac Not tested, I'm for the dirty way :) Ciao. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c64 --- Comment #64 from Tejun Heo <teheo@novell.com> 2010-07-02 16:28:29 UTC --- Alberto, can you please write to linux-ide@vger.kernel.org and cc tj@kernel.org and include the following information? * How much the load cycle increases per hour. * hdparm, smartctl -a and dmidecode output as attachments. Thanks. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c65 --- Comment #65 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2010-07-02 17:03:26 UTC --- Yes. I do not know exactly how much the count increases per hour, but I can provide an estimate. Doing the math roughly I had 1183 cycles in 2 days of life of the machine. Considering the machine has been on for about 8-10 hours per day, it is about 60 cycles per hours. P.S. dmidecode attached here too for completeness :) -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c66 --- Comment #66 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2010-07-02 17:04:08 UTC --- Created an attachment (id=373581) --> (http://bugzilla.novell.com/attachment.cgi?id=373581) dmidecode on Latitude E6410 with Seagate HD -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=545024 http://bugzilla.novell.com/show_bug.cgi?id=545024#c67 Alberto Passalacqua <albert.passalacqua@gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Version|RC 1 |Final --- Comment #67 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2010-07-17 21:24:28 UTC --- This did not make in to openSUSE 11.3 final, and the problem is still there. Quite interestingly it has been reported during the development stage! Moving to final... -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=545024 https://bugzilla.novell.com/show_bug.cgi?id=545024#c68 Alberto Passalacqua <albert.passalacqua@gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Component|Kernel |Kernel Product|openSUSE 11.3 |openSUSE 11.4 Target Milestone|RC 1 |Final --- Comment #68 from Alberto Passalacqua <albert.passalacqua@gmail.com> 2011-03-12 06:47:15 UTC --- Same problem on openSUSE 11.4, on the same Latitude E6410 discussed on comment 66. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=545024 https://bugzilla.novell.com/show_bug.cgi?id=545024#c69 Jeff Mahoney <jeffm@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |RESOLVED Resolution| |WONTFIX --- Comment #69 from Jeff Mahoney <jeffm@suse.com> 2012-08-02 11:59:15 EDT --- With the coming release of openSUSE 12.2, openSUSE kernel developers are focusing their efforts there. Reports against openSUSE 11.4 and prior will not get the attention needed to resolve them before openSUSE 12.2 is release and openSUSE 11.4 becomes unmaintained. Please re-test with openSUSE 12.1 or openSUSE RC2+ and re-open with an updated Product if you still encounter your issue. We apologize for this issue not getting the attention it deserves but we are focusing our resources in the area where they will have the most impact for our users. We're working hard to make openSUSE 12.2 the best openSUSE release yet! -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
participants (1)
-
bugzilla_noreply＠novell.com