There is this time skip during kernel boot:
Jul 19 14:07:02 deetabokkusu kernel: [ 2.874776] hub 1-1.4:1.0: USB hub found
Jul 19 14:07:02 deetabokkusu kernel: [ 2.877993] hub 1-1.4:1.0: 4 ports detected
Jul 19 14:07:02 deetabokkusu kernel: [ 94.080339] raid6: sse2x1 gen() 858 MB/s
Jul 19 14:07:02 deetabokkusu kernel: [ 94.128306] raid6: sse2x1 xor() 2371 MB/s
Jul 19 14:07:02 deetabokkusu kernel: [ 94.176324] raid6: sse2x2 gen() 1730 MB/s
Jul 19 14:07:02 deetabokkusu kernel: [ 94.224328] raid6: sse2x2 xor() 2816 MB/s
Jul 19 14:07:02 deetabokkusu kernel: [ 94.272285] raid6: sse2x4 gen() 2430 MB/s
Jul 19 14:07:02 deetabokkusu kernel: [ 94.320302] raid6: sse2x4 xor() 2333 MB/s
Jul 19 14:07:02 deetabokkusu kernel: [ 94.323121] raid6: using algorithm sse2x4 gen() 2430 MB/s
Jul 19 14:07:02 deetabokkusu kernel: [ 94.325949] raid6: .... xor() 2333 MB/s, rmw enabled
Jul 19 14:07:02 deetabokkusu kernel: [ 94.328762] raid6: using ssse3x2 recovery algorithm
Jul 19 14:07:02 deetabokkusu kernel: [ 94.335698] xor: measuring software checksum speed
Jul 19 14:07:02 deetabokkusu kernel: [ 94.376549] prefetch64-sse: 7668.000 MB/sec
Jul 19 14:07:02 deetabokkusu kernel: [ 94.416564] generic_sse: 6661.000 MB/sec
Jul 19 14:07:02 deetabokkusu kernel: [ 94.419385] xor: using function: prefetch64-sse (7668.000 MB/sec)
Jul 19 14:07:02 deetabokkusu kernel: [ 94.426733] async_tx: api initialized (async)
Jul 19 14:07:02 deetabokkusu kernel: [ 94.555496] Btrfs loaded, crc32c=crc32c-intel
systemd-analyze
returns:
Startup finished in 1min 34.896s (kernel) + 38.914s (userspace) = 2min 13.810s
graphical.target reached after 5.169s in userspace
systemd-analyze blame
returns:
25.913s apt-daily.service
9.181s apt-daily-upgrade.service
2.131s dev-sda2.device
1.325s snapd.service
729ms media-duraibu1.mount
717ms networkd-dispatcher.service
I am not running a RAID on the system. What do I have to change for the kernel to boot more quickly?
Edit:
What I am interested in is the kernel boot time of 1 min 34.896s. I know how to improve the userspace boot time of 38.914s with the information systemd-analyze blame
gave me.
Edit 2: Booting with two other LTS Kernels (v4.19.59 and v4.14.133) did not change the problem
systemd-analyze blame
. – Michal Przybylowicz Jul 19 '19 at 17:48system-analyze blame
only shows the time up take for userspace services, doesn't it? the highest uptake is "3.290s dev-sda2.device" -> that seems ok to me. the problem is the kernel startup time which is at 1min 35.095s as shown bysystemd-analyze
:Startup finished in 1min 35.095s (kernel) + 3.321s (userspace) = 1min 38.416s
– nucleus Jul 19 '19 at 17:57systemd-analyze blame
to find out what is taking so long. Note you posted same question on SuperUser: https://superuser.com/questions/1461801/ubuntu-kernel-boot-stuck-at-raid6-sse2x1-gen-858-mb-s Did you opt forbtrfs
file system instead of conventionalext4
file system? – WinEunuuchs2Unix Jul 20 '19 at 02:10systemd-analyze blame
to my post. But as I pointed out earlier it does not help me with my problem as it only shows me the time uptake of services that start during userspace startup. I am interested in the services that start during kernel startup which consume way more time. I'll look into changing the file system and booting with a different kernel version and post my results later! – nucleus Jul 20 '19 at 10:05