Bug 985076

Summary: Boot slow
Product: [Fedora] Fedora Reporter: Maarten <m.menheere>
Component: systemdAssignee: systemd-maint
Status: CLOSED DUPLICATE QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: 19CC: dennis, johannbg, lnykryn, mschmidt, msekleta, notting, plautrba, systemd-maint, vpavlin, zbyszek
Target Milestone: ---   
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2013-09-13 03:21:30 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Attachments:
Description Flags
systemd-analyse plot
none
boot.log file plymouth-wait-quit failure
none
boot with plymouth.enable=0
none
Boot with systemd.log_level=debug, output from journalctl -b
none
with debug not debugh
none
/etc/fstab
none
boot plot comment out
none
systemctl mask plymouth-start.service
none
systemctl mask accounts-daemon.service
none
both accounts daemon and firewall mask
none
all slow services masked none

Description Maarten 2013-07-16 17:54:24 UTC
Created attachment 774438 [details]
systemd-analyse plot

Description of problem:
Startup time is 1.31min very consistent. Disableling service does not change boot time. 

Version-Release number of selected component (if applicable):


How reproducible:


Steps to Reproduce:
1. Boot system
2.
3.

Actual results:


Expected results:


Additional info:

Comment 1 Jóhann B. Guðmundsson 2013-07-16 19:40:22 UTC
Not sure why this was moved against system since these are the slowdown components..

NetworkManager.service
restorecond.service
accounts-daemon.service
rtkit-daemons.service
plymouth-quit-wait.service
gdm.service
NetworkManager-wait-online.service

And disabling what services? 

( might be related to #985038 )

Comment 2 Maarten 2013-07-16 21:17:59 UTC
LVM, Firewalld, Avahi, chrony. They are all in the 20 seconds on systemd-analyse blame. Remove one. Reboot run blame again. The next one moves up in the 20 second range. (firewall is back on :) ) Included boot.log

Comment 3 Maarten 2013-07-16 21:20:13 UTC
Created attachment 774506 [details]
boot.log file plymouth-wait-quit failure

Comment 4 Maarten 2013-07-16 21:39:33 UTC
Created attachment 774509 [details]
boot with plymouth.enable=0

Comment 5 Maarten 2013-07-16 21:41:11 UTC
Boottime does not change. 

Startup finished in 1.201s (kernel) + 959ms (initrd) + 1min 31.927s (userspace) = 1min 34.088s

Comment 6 Michal Schmidt 2013-07-17 06:15:35 UTC
Please boot with "systemd.log_level=debug" on the kernel command line and then attach the output of the command "journalctl -b".

Comment 7 Maarten 2013-07-17 16:58:28 UTC
Created attachment 774861 [details]
Boot with systemd.log_level=debug, output from journalctl -b

Comment 8 Jóhann B. Guðmundsson 2013-07-17 17:09:00 UTC
That output does not contain debug output since you added "systemd.log_level=debugh" debugh instead of debug so you need to do it again and can you attach your /etc/fstab as well since there is most likely something wrong with it 

systemd[1]: Mounted /mnt/26CEDA80CEDA47A7.

jul 17 18:55:00 localhost.localdomain systemd[1]: Job dev-disk-by\x2duuid-48A00C08A00BFAE8.device/start timed out.
jul 17 18:55:00 localhost.localdomain systemd[1]: Timed out waiting for device dev-disk-by\x2duuid-48A00C08A00BFAE8.device.
jul 17 18:55:00 localhost.localdomain systemd[1]: Dependency failed for /mnt/48A00C08A00BFAE8.

What are you mounting here 

/mnt/26CEDA80CEDA47A7?

and here

/mnt/48A00C08A00BFAE8 < which is failing?

Comment 9 Maarten 2013-07-17 21:38:34 UTC
Created attachment 774991 [details]
with debug not debugh

Sorry, microsoft natural F10 is h when not in f-lock whish is special spell funcion.

Comment 10 Jóhann B. Guðmundsson 2013-07-17 22:49:22 UTC
Still missing your /etc/fstab

Comment 11 Maarten 2013-07-18 05:49:38 UTC
Created attachment 775108 [details]
/etc/fstab

Comment 12 Jóhann B. Guðmundsson 2013-07-18 17:07:37 UTC
What happens if you comment out those mnt entries there and or change them to use UUID= instead ( you can just run the blkid command to find the UUID entries )

Comment 13 Maarten 2013-07-18 18:50:22 UTC
Created attachment 775468 [details]
boot plot comment out

Looks a lot better now. 37s is ok.

Comment 14 Jóhann B. Guðmundsson 2013-07-18 18:57:00 UTC
37s is probably what people get out from our crappy defaults on rotating media. 

By disabling the stuff you dont actually need ( unless you need everything there ) you should be able to get down to 12s on rotating media and about <4s on SSD

What kind of hardware are you running on?

Comment 15 Maarten 2013-07-18 19:17:22 UTC
CPU: Intel(R) Core(TM)2 Duo E8400  @ 3.00GHz
Memory: 2GB
Motherboard: ASUS P5Q  (intel P45)
Videocard: [AMD/ATI] RV620 LE [Radeon HD 3450]
Harddrive: Seagate Barracuda 7200.12 500GB ( ST3500418AS ). 

Already tried to turn off as much as possible. 12s would be nice :). 

Still have this boot error message. 
[FAILED] Failed to start Wait for Plymouth Boot Screen to Quit.
See 'systemctl status plymouth-quit-wait.service' for details.

Looks like most time is spent on this in the boot plot.

Comment 16 Maarten 2013-07-18 20:19:25 UTC
Created attachment 775500 [details]
systemctl mask plymouth-start.service

Startup from 37s to 25s without plymouth. Seems like a bit much. Just benchmarking for fun here

Comment 17 Maarten 2013-07-19 06:04:14 UTC
Created attachment 775642 [details]
systemctl mask accounts-daemon.service

Comment 18 Maarten 2013-07-19 06:04:52 UTC
Created attachment 775643 [details]
both accounts daemon and firewall mask

Comment 19 Maarten 2013-07-19 06:05:37 UTC
Created attachment 775644 [details]
all slow services masked

Comment 20 Lennart Poettering 2013-09-13 03:21:30 UTC
Looks like it was the same as bug 983688

*** This bug has been marked as a duplicate of bug 983688 ***