How to fix slow startup?

Question

I am running a dual-boot desktop with Windows 10 and Ubuntu 16.04. While my windows startup is very fast, my Ubuntu takes too much time. To investigate, I ran dmesg and found some messages which I believe might be the cause. However, I am not sure if that is actually the case and if I can do something to speedup the boot

Here are the last lines of dmesg

[   14.635266] Adding 4881404k swap on /dev/sda5.  Priority:-1 extents:1 across:4881404k FS
[   22.877684] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[   22.884878] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[   22.901317] IPv6: ADDRCONF(NETDEV_UP): enp6s0: link is not ready
[   22.901971] atl1c 0000:06:00.0: atl1c: enp6s0 NIC Link is Up<1000 Mbps Full Duplex>
[   24.055406] IPv6: ADDRCONF(NETDEV_UP): wlp3s0: link is not ready
[   27.733484] audit_printk_skb: 42 callbacks suppressed
[   27.733487] audit: type=1400 audit(1495500079.496:26): apparmor="DENIED" operation="open" profile="/usr/sbin/mysqld" name="/proc/1080/status" pid=1080 comm="mysqld" requested_mask="r" denied_mask="r" fsuid=121 ouid=121
[   27.733510] audit: type=1400 audit(1495500079.496:27): apparmor="DENIED" operation="open" profile="/usr/sbin/mysqld" name="/sys/devices/system/node/" pid=1080 comm="mysqld" requested_mask="r" denied_mask="r" fsuid=121 ouid=0
[   27.733538] audit: type=1400 audit(1495500079.496:28): apparmor="DENIED" operation="open" profile="/usr/sbin/mysqld" name="/proc/1080/status" pid=1080 comm="mysqld" requested_mask="r" denied_mask="r" fsuid=121 ouid=121
[  304.593596] SGI XFS with ACLs, security attributes, realtime, no debug enabled
[  304.631262] JFS: nTxBlock = 8192, nTxLock = 65536
[  304.666374] ntfs: driver 2.1.32 [Flags: R/O MODULE].
[  304.704295] QNX4 filesystem 0.2.3 registered.
[  304.817377] raid6: sse2x1   gen()  7149 MB/s
[  304.885373] raid6: sse2x1   xor()  6892 MB/s
[  304.953370] raid6: sse2x2   gen() 11874 MB/s
[  305.021366] raid6: sse2x2   xor()  8411 MB/s
[  305.089361] raid6: sse2x4   gen() 13842 MB/s
[  305.157356] raid6: sse2x4   xor() 10249 MB/s
[  305.157357] raid6: using algorithm sse2x4 gen() 13842 MB/s
[  305.157358] raid6: .... xor() 10249 MB/s, rmw enabled
[  305.157359] raid6: using ssse3x2 recovery algorithm
[  305.173207] xor: automatically using best checksumming function:
[  305.209351]    avx       : 23160.000 MB/sec
[  305.252714] Btrfs loaded

Update Here is the output of systemd-analyze blame

     16.105s grub-common.service
     15.804s [email protected]
     14.856s ondemand.service
     14.080s networking.service
     14.057s apport.service
     13.611s mysql.service
     13.265s speech-dispatcher.service
     12.494s irqbalance.service
     10.698s sysstat.service
     10.641s lightdm.service
      7.857s dev-sda6.device
      5.599s ModemManager.service
      4.685s apparmor.service
      4.666s accounts-daemon.service
      3.626s NetworkManager.service
      2.811s redis-server.service
      2.670s systemd-logind.service
      2.583s upower.service
      2.506s gpu-manager.service
      2.442s systemd-user-sessions.service
      2.061s console-setup.service
      1.994s thermald.service
      1.756s systemd-tmpfiles-setup.service
      1.726s setvtrgb.service
      1.243s keyboard-setup.service
      1.194s systemd-udevd.service
      1.166s rsyslog.service
      1.054s plymouth-start.service
      1.026s systemd-tmpfiles-setup-dev.service
       902ms colord.service
       868ms avahi-daemon.service
       614ms systemd-modules-load.service
       578ms wpa_supplicant.service
       555ms systemd-rfkill.service
       537ms systemd-journald.service
       420ms sys-kernel-debug.mount
       420ms dev-mqueue.mount
       419ms dev-hugepages.mount
       348ms polkitd.service
       345ms systemd-update-utmp.service
       345ms systemd-timesyncd.service
       337ms apt-daily.service
       336ms systemd-sysctl.service
       333ms dns-clean.service
       311ms udisks2.service
       310ms plymouth-read-write.service
       301ms systemd-journal-flush.service
       276ms ufw.service
       276ms kmod-static-nodes.service
       214ms systemd-udev-trigger.service
       192ms resolvconf.service
       132ms systemd-random-seed.service
       126ms dev-disk-by\x2duuid-0b399021\x2dd995\x2d469b\x2d9027\x2d01183ad502e6.swap
       100ms systemd-remount-fs.service
        91ms pppd-dns.service
        36ms snapd.socket
        17ms [email protected]
         8ms snapd.autoimport.service
         7ms alsa-restore.service
         3ms ureadahead-stop.service
         3ms systemd-update-utmp-runlevel.service
         2ms rtkit-daemon.service
         2ms rc-local.service
         1ms sys-fs-fuse-connections.mount
       900us plymouth-quit-wait.service
       670us postgresql.service

Answer 1

The best way to really diagnose issues is to use a combination of calls from systemd-analyze. This tool not only helps you identify issues, but will give you visual representations as well.

1. Perform a basic diagnosis as shown below. Look at the most pervasive items and determine if you can't stop the autostart or modify some configurations

     systemd-analyze blame
      8.121s apt-daily.service
      7.658s NetworkManager-wait-online.service
      931ms docker.service
      710ms winbind.service
      695ms nmbd.service
      647ms samba-ad-dc.service
      543ms ModemManager.service
   

2. If the listing above doesn't explain it enough, then the next best option is to using the plotting tool. This will output the entire boot process into an svg file. Yep then you can visually see the issues! You simply type systemd-analyze plot > test.svg

3. Open the test.svg and have a look at the various services that are plotting a larger amount of time. Then remove them from autostart or determine if there are configurations that can be made to fix them

4. Reboot the machine and run Step 2 again to see if there are other remaining issues