What is the best way to run ntpdate at reboot, only after network is ready
I'm using a BeagleBone, and since it has no built in RTC and battery back up, it loses the date on every reboot. I can easily set the date with the command:
/usr/bin/ntpdate -b -s -u pool.ntp.org
But if the power goes out and back on for the house for example, then the time is lost. The solution that comes with the latest beaglebone Angstrom linux distribution is to put a crontab line in that updates the time every half hour. But I would prefer to just run the command once on powerup.
I tried putting this command listed above in crontab with the @reboot line, but that I believe ran before network was configured, or something else failed since it didn't get me the right time when I pulled the power for 5 minutes and put it back to the beaglebone.
Is there some way to use ifconfig or something like that to run a script from init.d only after network is available?
Answer
opkg install ntp-systemd
systemctl enable ntpdate.service
systemctl enable ntpd.service
Edit /etc/ntp.conf and comment the following lines (no fallback on an hardware clock that doesn't exist and because the ntpdate service use the "ntpd -q" command)
#server 127.127.1.0
#fudge 127.127.1.0 stratum 14
Two services are installed:
/lib/systemd/system/ntpd.service:
[Unit]
Description=Network Time Service
After=network.target
[Service]
Type=forking
PIDFile=/run/ntpd.pid
ExecStart=/usr/bin/ntpd -p /run/ntpd.pid
/lib/systemd/system/ntpdate.service:
[Unit]
Description=Network Time Service (one-shot ntpdate mode)
Before=ntpd.service
[Service]
Type=oneshot
ExecStart=/usr/bin/ntpd -q -g -x
RemainAfterExit=yes
ntpd is started after the network is up (After=network.target) so the date should be continuously synchronized. BUT has explained in the ntpd man page:
Most operating systems and hardware of today incorporate a time-of-year (TOY) chip to maintain the time during periods when the power is off. When the machine is booted, the chip is used to initialize the operating system time. After the machine has synchronized to a NTP server, the operating system corrects the chip from time to time. In case there is no TOY chip or for some reason its time is more than 1000s from the server time, ntpd assumes something must be terribly wrong and the only reliable action is for the operator to intervene and set the clock by hand. This causes ntpd to exit with a panic message to the system log. The -g option overrides this check and the clock will be set to the server time regardless of the chip time. However, and to protect against broken hardware, such as when the CMOS battery fails or the clock counter becomes defective, once the clock has been set, an error greater than 1000s will cause ntpd to exit anyway.
So we need to set the date before starting ntpd and this is done by the ntpdate service by executing "ntpd -q -g -x" before starting ntpd.service.
From ntpd man page:
-q Exit the ntpd just after the first time the clock is set. This behavior mimics that of the ntpdate program, which is to be retired. The -g and -x options can be used with this option. Note: The kernel time discipline is disabled with this option.
Another service installed on the Beaglebone interact with the date/time
timestamp.service
[Unit]
Description=Timestamping service
ConditionPathExists=/etc/timestamp
After=remount-rootfs.service
[Service]
RemainAfterExit=yes
ExecStart=/usr/bin/load-timestamp.sh
ExecStop=/usr/bin/load-timestamp.sh --save
This service store the current timestamp in /etc/timestamp when it's stopped and set the date from that timestamp when it's started. So if ntpd isn't installed, the date set manually and the beaglebone rebooted, the date is only behind by the boot duration.