New: journal activity on inactive partition causes inactive harddrive spinup
A message that is not in the archive (sent off-list or not yet received): [email protected]/
(switched to email. Please respond via emailed reply-to-all, not via the bugzilla web interface).
Quoted 18 lines · On Sun, 14 Dec 2008 11:39:34 -0800 (PST) [email protected] wrote:
http://bugzilla.kernel.org/show_bug.cgi?id=12224
Summary: journal activity on inactive partition causes inactive
harddrive spinup
Product: File System
Version: 2.5
KernelVersion: 2.6.27.6
Platform: All
OS/Version: Linux
Tree: Mainline
Status: NEW
Severity: low
Priority: P1
Component: ext3
AssignedTo: [email protected]
ReportedBy: [email protected]
Quoted 2 lines
Latest working kernel version: 2.6.27.5
Earliest failing kernel version: 2.6.27.6
Thanks for being this specific.
Quoted 10 lines
Distribution: Debian stable with custom compiled vanilla kernel
Problem Description:
Activity on one partition (hda1) causes kjournald to also write to another
partition (sda1). This in turn causes the harddrive which contains sda1 to spinup or not spindown.
I've activated laptop-mode using laptop-mode-tools. hda1 and sda1 are both
formatted with ext3.
On unaffected kernels (e.g. 2.6.27.5) btrace produces no output. On
affected kernels (e.g. 2.6.27.6, 2.6.27-rc7) btrace produces output as below.
I think I've seen other reports of this. Maybe it was you.
Quoted 15 lines
/dev/hda1on / type ext3 (rw,noatime,errors=remount-ro,commit=1209600) tmpfs on/lib/init/rwtype tmpfs (rw,nosuid,mode=0755)
proc on /proc type proc (rw,noexec,nosuid,nodev)
sysfs on /sys type sysfs (rw,noexec,nosuid,nodev)
udev on /dev type tmpfs (rw,mode=0755)
tmpfs on/dev/shmtype tmpfs (rw,nosuid,nodev)
devpts on/dev/ptstype devpts (rw,noexec,nosuid,gid=5,mode=620) debugfs on/sys/kernel/debugtype debugfs (rw)Steps to reproduce:
1. start "btrace/dev/sda"
2. begin compiling kernel
3. within seconds btrace begins output similar to below.
Quoted 4 lines
Thanks for all your work! Sorry if I've misplaced blame for this undesirable change in behavior.
Using 2.6.27.8, btrace of
/dev/sdawhile compiling on/dev/hda1
OK, that's a huge mess because bugzilla went and wordwrapped it. Let me fix it:
Quoted 8 lines
Dec 9 11:15:32 localhost logger: 8,0 0 1 0.000000000 1109 A WS 14375 + 8 <- (8,1) 14312 Dec 9 11:15:32 localhost logger: 8,0 0 2 0.000006705 1109 Q WS 14375 + 8 [kjournald] Dec 9 11:15:32 localhost logger: 8,0 0 3 0.000043581 1109 G WS 14375 + 8 [kjournald] Dec 9 11:15:32 localhost logger: 8,0 0 4 0.000065651 1109 P N [kjournald] Dec 9 11:15:32 localhost logger: 8,0 0 5 0.000076267 1109 I W 14375 + 8 [kjournald] Dec 9 11:15:32 localhost logger: 8,0 0 6 0.000300318 1109 D W 14375 + 8 [kjournald] Dec 9 11:15:32 localhost logger: 8,0 0 7 0.000423518 1109 U N [kjournald] 1 Dec 9 11:15:32 localhost logger: 8,0 0 8 0.010028373 3 C W 14375 + 8 [0]
Quoted 8 lines
Dec 9 11:15:35 localhost logger: 8,0 0 9 3.219049689 1109 A WS 14383 + 8 <- (8,1) 14320 Dec 9 11:15:35 localhost logger: 8,0 0 10 3.219058350 1109 Q WS 14383 + 8 [kjournald] Dec 9 11:15:35 localhost logger: 8,0 0 11 3.219095505 1109 G WS 14383 + 8 [kjournald] Dec 9 11:15:35 localhost logger: 8,0 0 12 3.219115061 1109 P N [kjournald] Dec 9 11:15:35 localhost logger: 8,0 0 13 3.219127074 1109 I W 14383 + 8 [kjournald] Dec 9 11:15:35 localhost logger: 8,0 0 14 3.219344699 1109 D W 14383 + 8 [kjournald] Dec 9 11:15:35 localhost logger: 8,0 0 15 3.219543607 1109 U N [kjournald] 1 Dec 9 11:15:35 localhost logger: 8,0 0 16 3.229477275 3 C W 14383 + 8 [0]
I'm not familar with btrace. What are we looking at here?
The only likely commit I see in ftp://ftp.kernel.org/pub/linux/kernel/v2.6/ChangeLog-2.6.27.6 is "ext3:
wait on all pending commits in ext3_sync_fs", but I don't see how it could cause this.