| 1 |
Am 09.03.2014 18:39, schrieb Canek Peláez Valdés: |
| 2 |
|
| 3 |
> Something is wrong here; unless you are booting a 386 machine, there |
| 4 |
> is no way it should take a minute and a half to boot. And even with a |
| 5 |
> 386 I would be suspicious. |
| 6 |
|
| 7 |
No, actually it is an Intel i5-4670K with 8 GB of RAM. |
| 8 |
|
| 9 |
> Something is seriously wrong with systemd-tmpfiles-clean.service; why |
| 10 |
> it takes 1:07 minutes to run? Do you have /tmp as a tmpfs? |
| 11 |
|
| 12 |
Yes, at least according to mount, it is. |
| 13 |
|
| 14 |
mount | grep tmpfs |
| 15 |
devtmpfs on /dev type devtmpfs |
| 16 |
(rw,relatime,size=238864k,nr_inodes=59716,mode=755) |
| 17 |
tmpfs on /dev/shm type tmpfs (rw,nosuid,nodev) |
| 18 |
tmpfs on /run type tmpfs (rw,nosuid,nodev,mode=755) |
| 19 |
tmpfs on /sys/fs/cgroup type tmpfs (rw,nosuid,nodev,noexec,mode=755) |
| 20 |
tmpfs on /tmp type tmpfs (rw) |
| 21 |
|
| 22 |
|
| 23 |
By the way my actual blame does tell: |
| 24 |
|
| 25 |
systemd-analyze blame |
| 26 |
6.087s NetworkManager.service |
| 27 |
5.310s alsa-restore.service |
| 28 |
4.226s systemd-logind.service |
| 29 |
3.660s lightdm.service |
| 30 |
2.581s systemd-vconsole-setup.service |
| 31 |
688ms polkit.service |
| 32 |
479ms systemd-user-sessions.service |
| 33 |
413ms kmod-static-nodes.service |
| 34 |
381ms systemd-udev-trigger.service |
| 35 |
358ms user@×.service |
| 36 |
352ms systemd-tmpfiles-setup-dev.service |
| 37 |
274ms tmp.mount |
| 38 |
265ms systemd-journal-flush.service |
| 39 |
246ms systemd-sysctl.service |
| 40 |
235ms systemd-random-seed.service |
| 41 |
205ms upower.service |
| 42 |
205ms udisks2.service |
| 43 |
197ms systemd-udevd.service |
| 44 |
195ms systemd-tmpfiles-setup.service |
| 45 |
183ms systemd-fsck-root.service |
| 46 |
163ms systemd-remount-fs.service |
| 47 |
126ms systemd-update-utmp.service |
| 48 |
125ms sys-fs-fuse-connections.mount |
| 49 |
53ms wpa_supplicant.service |
| 50 |
50ms user@××××.service |
| 51 |
50ms accounts-daemon.service |
| 52 |
|
| 53 |
Actual systemd-analyze: |
| 54 |
|
| 55 |
Startup finished in 584542y 2w 2d 20h 1min 42.032s (loader) + 1.540s |
| 56 |
(kernel) + 11.028s (userspace) = 12.569s |
| 57 |
|
| 58 |
Actual critical chain: |
| 59 |
|
| 60 |
graphical.target @11.028s |
| 61 |
└─multi-user.target @11.028s |
| 62 |
└─NetworkManager.service @4.940s +6.087s |
| 63 |
└─basic.target @4.939s |
| 64 |
└─timers.target @4.721s |
| 65 |
└─systemd-tmpfiles-clean.timer @4.721s |
| 66 |
└─sysinit.target @4.489s |
| 67 |
└─systemd-vconsole-setup.service @1.907s +2.581s |
| 68 |
└─systemd-journald.socket @1.660s |
| 69 |
└─-.mount @1.660s |
| 70 |
└─system.slice @2.030s |
| 71 |
└─-.slice @2.030s |
| 72 |
|
| 73 |
> Could you run "systemd-analyze critical-chain systemd-tmpfiles-clean.service"? |
| 74 |
|
| 75 |
Sure, here it is: |
| 76 |
|
| 77 |
└─systemd-tmpfiles-clean.timer @4.721s |
| 78 |
└─sysinit.target @4.489s |
| 79 |
└─systemd-vconsole-setup.service @1.907s +2.581s |
| 80 |
└─systemd-journald.socket @1.660s |
| 81 |
└─-.mount @1.660s |
| 82 |
└─system.slice @2.030s |
| 83 |
└─-.slice @2.030s |
| 84 |
|
| 85 |
|
| 86 |
> In your critical-chain systemd-tmpfiles-clean.service was not included |
| 87 |
> (only systemd-tmpfiles-clean.timer). From blame, I think that's the |
| 88 |
> obvious offender. Again, do you have /tmp as a tmpfs? What do you have |
| 89 |
> in /etc/tmpfiles.d? |
| 90 |
|
| 91 |
/etc/tmpfiles.d is empty. |
| 92 |
|
| 93 |
> Notice that systemd-tmpfiles-clean.service takes almost no time; here |
| 94 |
> it's its critical chain: |
| 95 |
|
| 96 |
Yes, I see, so makes me wonder. |
| 97 |
|
| 98 |
BTW, my fstab: |
| 99 |
|
| 100 |
/dev/sda1 /boot ext2 noauto,noatime 0 0 |
| 101 |
/dev/sda2 / xfs noatime,nodiratime 0 0 |
| 102 |
|
| 103 |
|
| 104 |
Thanks. |
Archives