Re: GRUB really slow to boot
On Sun, 19 Dec 2021, Greg Wooledge wrote:
On Sun, Dec 19, 2021 at 07:19:40AM +0000, Tim Woodall wrote:
Check if the kernel log jumps from 1/1/70 to today as it boots. That
would point to the RTC being bad when the kernel first starts.
Not sure which log I'd need to look at for this information. dmesg only
reports time in relative seconds from the kernel's boot.
This is the sort of thing I meant (this is a rpi which doesn't have an
RTC at all so this is expected) - I've just rebooted to show this:
This is from daemon.log but any log should do.
Dec 19 14:52:42 rpi4-minimal haveged: haveged: Stopping due to signal 15
Jan 1 00:00:23 rpi4-minimal haveged: haveged starting up
...
Jan 1 00:00:23 rpi4-minimal ntpd[1522]: Listening on routing socket on fd #24 for interface updates
Jan 1 00:00:23 rpi4-minimal ntpd[1522]: kernel reports TIME_ERROR: 0x41: Clock Unsynchronized
Jan 1 00:00:23 rpi4-minimal ntpd[1522]: kernel reports TIME_ERROR: 0x41: Clock Unsynchronized
Jan 1 01:00:27 rpi4-minimal dbus-daemon[1664]: [session uid=103 pid=1659] Activating service name='org.xfce.Xfconf' requested by ':1.1' (uid=103 pid=1645 comm="x-window-manager ")
Jan 1 01:00:27 rpi4-minimal dbus-daemon[1664]: [session uid=103 pid=1659] Successfully activated service 'org.xfce.Xfconf'
Dec 19 14:53:40 rpi4-minimal dhclient[1176]: DHCPDISCOVER on eth0 to 255.255.255.255 port 67 interval 6
The clock corrects itself when ntp starts. I'm not exactly clear why the
clock jumps by an hour first but I don't really care...
/var/log/kern.log.1 shows this:
Dec 18 09:58:48 unicorn kernel: [6031224.812397] xor: automatically using best c
hecksumming function avx
Dec 18 09:58:48 unicorn kernel: [6031224.944868] Btrfs loaded, crc32c=crc32c-int
el
Dec 18 11:17:28 unicorn kernel: [ 0.000000] microcode: microcode updated earl
y to revision 0xea, date = 2021-01-05
Dec 18 11:17:28 unicorn kernel: [ 0.000000] Linux version 5.10.0-10-amd64 (de
bian-kernel@lists.debian.org) (gcc-10 (Debian 10.2.1-6) 10.2.1 20210110, GNU ld
(GNU Binutils for Debian) 2.35.2) #1 SMP Debian 5.10.84-1 (2021-12-08)
Of course, the human-readable timestamps on the left come from syslog,
which is a userspace process that is launched after the clock is
initialized from whatever sources (not sure exactly when NTP kicks in).
Looks to me at first glance that the clock was right - so scratch my
guess that it could be a dead battery on the motherboard.
Tim.
Reply to: