升级到 Ubuntu 20.04 后，我的系统启动非常慢。
systemd-analyze
Startup finished in 5.731s (firmware) + 4.338s (loader) + 5.616s (kernel) + 1min 25.744s (userspace) = 1min 41.430s
graphical.target reached after 1min 25.711s in userspace
...
systemd-analyze blame
35.978s plymouth-quit-wait.service >
24.686s dev-sda6.device >
21.190s systemd-journal-flush.service >
17.983s udisks2.service >
15.719s accounts-daemon.service >
13.912s gpu-manager.service >
13.267s NetworkManager.service >
12.942s avahi-daemon.service >
12.580s polkit.service >
12.294s switcheroo-control.service >
12.285s thermald.service >
12.282s wpa_supplicant.service >
12.282s systemd-logind.service >
10.601s dev-loop3.device >
10.456s dev-loop7.device >
10.435s dev-loop0.device >
10.030s ModemManager.service >
9.727s dev-loop4.device >
9.282s dev-loop10.device >
9.269s dev-loop8.device >
9.052s dev-loop5.device >
9.040s dev-loop6.device >
9.024s dev-loop1.device
...
systemd-analyze critical-chain
The time when unit became active or started is printed after the "@" character.
The time the unit took to start is printed after the "+" character.
graphical.target @1min 25.711s
└─multi-user.target @1min 25.710s
└─ModemManager.service @49.195s +10.030s
└─polkit.service @36.607s +12.580s
└─basic.target @35.803s
└─sockets.target @35.803s
└─uuidd.socket @35.803s
└─sysinit.target @35.461s
└─snapd.apparmor.service @34.460s +1.000s
└─apparmor.service @32.093s +2.364s
└─local-fs.target @32.092s
└─run-user-129-gvfs.mount @1min 11.784s
└─run-user-129.mount @1min 9.908s
└─swap.target @30.318s
└─dev-disk-by\x2duuid-2b54997c\x2d1642\x2d411d\x2d9d0>
└─dev-disk-by\x2duuid-2b54997c\x2d1642\x2d411d\x2d9>
答案1
https://wiki.archlinux.org/index.php/Improving_performance/Boot_process
基本上 systemd 很乱，所以你可以改进它，但改进的程度不大。
答案2
遇到类似问题。全新安装运行良好……试图运行 Clonezilla，这是另一个问题，在关机期间过早拉出安装介质。我能够使用启动修复来拯救系统，但启动到 Ubuntu 20.02 现在需要一分半钟。