Very Slow Boot in F16

Raphy · #1 26th February 2012, 11:15 AM

Hello,

Since some random updates, my computer is very slow to boot (about 5 minutes). The plymouth logo is slow loading during 5 minutes, until the SLIM manager start.

Here is my bootchart: http://i41.tinypic.com/2eupqhc.png

unfortunately I can't understand where the problem is. I disabled the icsi services but still nothing.

Thank for your help
Raphael

sea · #2 26th February 2012, 01:35 PM

Sorry cant recognized anything on the graphic.

Could you provide the output of either

Code:

systemd-analyze blame
systemd-analyze

And copy paste the output into [ code] ... [/ code] tags, OR

Code:

systemd-analyze blame > ~/blame.txt
systemd-analyze >> ~/blame.txt

And attach blame.txt

Only disable iscsi service if you have no scsi devices attached to your environment.
As you are directing towards an harddrive issue, please provide also the following output:

Code:

lsblk
blkid

Raphy · #3 26th February 2012, 05:33 PM

Quote:

Originally Posted by sea

Sorry cant recognized anything on the graphic.

Could you provide the output of either

Code:

systemd-analyze blame
systemd-analyze

And copy paste the output into [ code] ... [/ code] tags, OR

Code:

systemd-analyze blame > ~/blame.txt
systemd-analyze >> ~/blame.txt

And attach blame.txt

Only disable iscsi service if you have no scsi devices attached to your environment.
As you are directing towards an harddrive issue, please provide also the following output:

Code:

lsblk
blkid

Hello, thanks for your reply:

Here's the output of systemd-analyse -blame && systemd-analyse:

Code:

  1149ms rsyslog.service
  1070ms autofs.service
   850ms vmware-workstation-server.service
   823ms vmware.service
   589ms wicd.service
   464ms smb.service
   455ms dkms_autoinstaller.service
   436ms livesys.service
   424ms systemd-logind.service
   410ms console-kit-log-system-start.service
   395ms udev-settle.service
   383ms auditd.service
   377ms avahi-daemon.service
   324ms abrt-ccpp.service
   320ms dbus.service
   320ms mcelog.service
   317ms abrt-vmcore.service
   310ms sshd-keygen.service
   302ms irqbalance.service
   276ms vboxdrv.service
   260ms systemd-readahead-collect.service
   227ms systemd-readahead-replay.service
   224ms systemd-binfmt.service
   192ms fcoe.service
   167ms media.mount
   164ms fedora-storage-init.service
   155ms bootchart.service
   150ms udev-trigger.service
   138ms mdmonitor-takeover.service
   138ms var-lib-nfs-rpc_pipefs.mount
   120ms remount-rootfs.service
   117ms systemd-remount-api-vfs.service
   117ms sys-kernel-security.mount
   108ms systemd-vconsole-setup.service
   107ms fedora-readonly.service
   102ms nfs-idmap.service
   101ms systemd-sysctl.service
    95ms sys-kernel-debug.mount
    81ms fedora-storage-init-late.service
    81ms nfs-lock.service
    74ms dev-mqueue.mount
    73ms rpcbind.service
    71ms systemd-tmpfiles-setup.service
    68ms chronyd.service
    54ms dev-hugepages.mount
    52ms console-kit-daemon.service
    46ms sys-kernel-config.mount
    45ms sandbox.service
    41ms lvm2-monitor.service
    41ms netfs.service
    39ms fedora-wait-storage.service
    38ms udev.service
    37ms livesys-late.service
    37ms vmware-USBArbitrator.service
    34ms vboxweb-service.service
    32ms vboxballoonctrl-service.service
    31ms home.mount
    30ms boot.mount
    28ms systemd-user-sessions.service
    15ms sys-bus-usb-drivers.mount
    12ms proc-sys-fs-binfmt_misc.mount
     4ms rtkit-daemon.service
     0ms sys-fs-fuse-connections.mount
Startup finished in 1692ms (kernel) + 2015ms (initramfs) + 306202ms (userspace) = 309910ms

As for the blkid command:

Code:

/dev/sda2: UUID="54d58a9b-1375-45cd-8480-cf70e4634e24" TYPE="ext4" 
/dev/sda3: UUID="de0553b8-1bbf-40a1-9f48-64fe2ff1b7d0" TYPE="swap" 
/dev/sda4: LABEL="_Fedora-16-x86_6" UUID="91ac9dc2-2d6a-44b7-9c5a-740093597180" TYPE="ext4" 
/dev/sda5: UUID="a6777649-6002-45ff-a7aa-3d6aee88f573" TYPE="ext4"

And the lsblk command:

Code:

NAME   MAJ:MIN RM   SIZE RO TYPE MOUNTPOINT
sda      8:0    0 119.2G  0 disk 
├─sda1   8:1    0     1M  0 part 
├─sda2   8:2    0   500M  0 part /boot
├─sda3   8:3    0   9.7G  0 part [SWAP]
├─sda4   8:4    0    50G  0 part /
└─sda5   8:5    0    59G  0 part /home
sr0     11:0    1  1024M  0 rom

Thanks again for your help
Raphy

sea · #4 26th February 2012, 06:49 PM

Ok for the optics at first:

This step is optional

Code:

seasc set os rh plymouth

Install the all plymouth files, and after change to either details or text.
As soon the boot time is acceptable, change to what ever you want (my preference is solar).

If you neither use: (to me, it looks like you dont)
* raid
* scsi
* lvm
* livesystem (NOT an installation on a usb. A livemedia)

If you do use any of the above, the according service should be removed from the list variable below.
You can run this in a terminal:

Code:

list="
fedora-storage-init
fedora-storage-init-late
fedora-wait-storage
mdmonitor-takeover
livesys
livesys-late
lvm2-monitor
"
su -c "for s in $list ; do systemctl disable $s.service;done"

After this step, reboot and post the output of:

Code:

systemd-analyze

Raphy · #5 26th February 2012, 08:29 PM

Thank you, I disabled all the services you mentionned, and now boot time is as short as ever.

Unfortunately now "systemd-analyze" fail with this error:

Code:

Traceback (most recent call last):
  File /usr/bin/systemd-analyze, line 82, in <module>
    initrd_time, start_time, finish_time = acquire_start_time()
  File /usr/bin/systemd-analyze, line 35, in acquire_start_time
    assert startup_time <= finish_time
AssertionError

Every thing all seems to be working fine though.

sea · #6 26th February 2012, 08:44 PM

Try to enable mdmonitor-takeover.service again, and see if boot time remains.

Code:

# systemctl enable mdmonitor-takeover.service

Raphy · #7 26th February 2012, 08:48 PM

Boot time remains. But systemd-analyse still fails with the same error.

sea · #8 26th February 2012, 08:54 PM

Ok, as i was guessing the mdmonitor, disable it again.
Next guess would be to enable fedora-wait-storage.

If systemd-analyze still doesnt work, i would not know which other service it could be.
I'm not that experienced enough with services, these 2 seemed to match best imho.

EDIT:
You have 2 options:
a) toggle all services of the list on and off, reboot after each one...
b) Most brute way to fix it would be

Code:

sudo yum reinstall systemd

Raphy · #9 26th February 2012, 09:04 PM

I can't enable this service again:
"Warning: unit files do not carry install information. No operation executed
"

I don't know how the other services could help with this error.

But you know, everything seems to be working fine, so thank you for your help already. Thank you very much

AlexDudko · #10 26th February 2012, 09:08 PM

I've got perhaps the same issue. Here's my systemd-analyse output:

Quote:

6706ms iscsid.service
5171ms netfs.service
4903ms NetworkManager.service
4284ms udev-settle.service
3937ms bluetooth.service
3354ms avahi-daemon.service
3084ms systemd-logind.service
2039ms fedora-loadmodules.service
1458ms systemd-binfmt.service
1401ms systemd-vconsole-setup.service
1105ms proc-sys-fs-binfmt_misc.mount
1054ms ip6tables.service
1017ms fedora-readonly.service
1017ms console-kit-log-system-start.service
909ms livesys.service
817ms iptables.service
793ms gpm.service
783ms mdmonitor.service
773ms remount-rootfs.service
749ms var-lock.mount
738ms var-run.mount
729ms media.mount
723ms systemd-remount-api-vfs.service
702ms irqbalance.service
689ms dev-mqueue.mount
620ms dev-hugepages.mount
611ms sys-kernel-config.mount
601ms sys-kernel-security.mount
593ms sys-kernel-debug.mount
546ms mcelog.service
535ms abrt-vmcore.service
498ms systemd-tmpfiles-setup.service
436ms udev-trigger.service
418ms fedora-storage-init.service
376ms udev.service
346ms mdmonitor-takeover.service
331ms rsyslog.service
279ms dbus.service
261ms sshd-keygen.service
241ms rc-local.service
130ms accounts-daemon.service
110ms dev-sda2.swap
109ms sm-client.service
98ms iscsi.service
93ms console-kit-daemon.service
84ms fedora-storage-init-late.service
84ms sendmail.service
63ms lvm2-monitor.service
60ms fedora-wait-storage.service
56ms systemd-sysctl.service
36ms systemd-user-sessions.service
20ms rtkit-daemon.service
18ms livesys-late.service
0ms sys-fs-fuse-connections.mount
Startup finished in 3314ms (kernel) + 4079ms (initramfs) + 76714ms (userspace) = 84109ms

sea · #11 26th February 2012, 11:21 PM

Just sharing my situation here, as i have no idea how to help further by any suggestions.

On my netbook, i have a file called scservices.
The harddisk is shown as ATA, however with lsscsi (i had to install it first) i get this output:

Code:

[0:0:0:0]    disk    ATA      Hitachi HTS54323 ES2O  /dev/sda

As it is a netbook, there is no raid possible, and i dont use lvm.
I do have a NAS, but most of the times it is offline and if its online i use the browser to access it.

The script looks like:

PHP Code:


			
#!/bin/bash
# |
# | Variables
    listS="fedora-readonly
        fedora-storage-init
        fedora-storage-init-late
        fedora-wait-storage
        ip6tables
        livesys
        livesys-late
        lvm2-monitor
        mdmonitor-takeover
        netfs
        sendmail
        "
# |
# | Display
    for s in $listS; do sudo systemctl disable $s.service;done
    for s in $listS;do sudo ln -sf /dev/zero /etc/systemd/system/$s.service;done
    
    sudo systemctl disable dev-hugepages.mount

While the boot times are:

Code:

systemd-analyze blame  && systemd-analyze
  4461ms systemd-vconsole-setup.service
  3064ms udev-trigger.service
  2731ms remount-rootfs.service
  2021ms media.mount
  1998ms sys-kernel-debug.mount
  1953ms sys-kernel-security.mount
  1909ms dev-hugepages.mount
  1872ms udev.service
  1860ms dev-mqueue.mount
  1645ms systemd-remount-api-vfs.service
  1432ms autofs.service
  1101ms NetworkManager.service
   999ms systemd-sysctl.service
   828ms rsyslog.service
   708ms systemd-logind.service
   705ms console-kit-log-system-start.service
   683ms .s.mount
   597ms iptables.service
   545ms avahi-daemon.service
   532ms auditd.service
   530ms systemd-readahead-replay.service
   480ms systemd-readahead-collect.service
   442ms acpid.service
   414ms sys-kernel-config.mount
   408ms irqbalance.service
   392ms sshd-keygen.service
   360ms dbus.service
   358ms systemd-tmpfiles-setup.service
   310ms systemd-user-sessions.service
   153ms boot.mount
   136ms home.mount
    92ms console-kit-daemon.service
    35ms sys-fs-fuse-connections.mount
    23ms rtkit-daemon.service
Startup finished in 1718ms (kernel) + 5385ms (initramfs) + 14277ms (userspace) = 21380ms

fanisatt · #12 3rd April 2012, 07:12 PM

I have always connected a TV BOX (USB 2.0) to my computer.
I am working with Linux and Windows XP on 2 separate hard disks (I can choose on boot my OS).
The TV BOX is working very good on Windows XP.
My Linux OS is the Fedora 16 KDE x86_64.
Fedora could not recognize the tv box but I had not any other problems with the OS.
Recently I got the official fedora updates (new kernel 3.3.0-8).
This kernel could recognise the usb tv box but, I could not find a way to use it.

Anyway , I noticed some unusual problems with the new kernel :
- Τhe boot time was very longer.
- Very often I could not use my keyboard and my mouse ον start up. I could not press any key to login..... (It is a usb keyboard - mouse). I had to restart the PC again and again....
- There was a persistent kernel bug reported to me very often by the ABRT immediately after login.
- During shutting off there was a bug (CPU #1 stall 22 sec). I had to press the power button of the PC steadily on order to shut off.

So, I decided to disconnect the tvbox of the usb port .
After this, all the above problems were disappeared.
I try to help.
Thanks a lot.