Network Time Protocol daemon: Difference between revisions

Latest revision as of 09:15, 27 September 2024

Network Time Protocol is the most common method to synchronize the software clock of a GNU/Linux system with internet time servers. It is designed to mitigate the effects of variable network latency and can usually maintain time to within tens of milliseconds over the public Internet. The accuracy on local area networks is even better, up to one millisecond.

The NTP Project provides a reference implementation of the protocol called simply NTP. This article further describes how to set up and run the NTP daemon, both as a client and as a server.

See System time#Time synchronization for other NTP implementations.

Installation

Install the ntp package. By default, ntpd works in client mode without further configuration. You can skip to #Usage if you want to use the Arch Linux default configuration file for it. For server configuration, see #NTP server mode.

Configuration

The main daemon is ntpd, which is configured in /etc/ntp.conf. Refer to ntp.conf(5) for detail.

Connection to NTP servers

NTP servers are classified in a hierarchical system with many levels called strata: the devices which are considered independent time sources are classified as stratum 0 sources; the servers directly connected to stratum 0 devices are classified as stratum 1 sources; servers connected to stratum 1 sources are then classified as stratum 2 sources and so on.

It has to be understood that a server's stratum cannot be taken as an indication of its accuracy or reliability. Typically, stratum 2 servers are used for general synchronization purposes: if you do not already know the servers you are going to connect to, you should choose a server pool close to your location from the pool.ntp.org servers (alternative link).

Since ntp version 4.2.7.p465-2, Arch Linux uses its own default vendor pool of NTP servers provided by the NTP Pool Project (see FS#41700). Modify those to suit your needs, e.g. if you want to use your country's servers with an option:

/etc/ntp.conf

server 0.fr.pool.ntp.org iburst
server 1.fr.pool.ntp.org iburst
server 2.fr.pool.ntp.org iburst
server 3.fr.pool.ntp.org iburst

The iburst option is recommended, and sends a burst of packets only if it cannot obtain a connection with the first attempt. The burst option always does this, even on the first attempt, and should never be used without explicit permission and may result in blacklisting.

Leap seconds file

In order for the system to be able to provide the International Atomic Time to an application that requests it, the list of leap seconds must be loaded. The list is part of the tzdata package and can be loaded by adding the following line to the NTP configuration file:

leapfile /usr/share/zoneinfo/leap-seconds.list

NTP server mode

If setting up an NTP server, check that you have orphan mode enabled, so that, in case it loses internet access, it will continue serving time to the network; enable orphan mode using the tos configuration parameter (you can set up to stratum 15) so that it will never be used unless internet access is lost:

tos orphan 15

Next, define the rules that will allow clients to connect to your service (localhost is considered a client too) using the restrict command; you should already have a line like this in your file:

restrict default nomodify nopeer noquery

This restricts everyone from modifying anything and prevents everyone from querying the status of your time server: nomodify prevents reconfiguring ntpd (with ntpq or ntpdc), and noquery is important to prevent dumping status data from ntpd (also with ntpq or ntpdc).

You can also add other options:

restrict default kod limited nomodify notrap nopeer noquery

Note: This still allows other people to query your time server. You need to add noserve to stop serving time. It will also block time synchronization since it blocks all packets except ntpq and ntpdc queries.

If you want to change any of these, see the full docs for the "restrict" option in ntp.conf(5), the detailed ntp instructions and #Usage.

Following this line, you need to tell ntpd what to allow through into your server; the following line is enough if you are not configuring an NTP server:

restrict 127.0.0.1

If you want to force DNS resolution to the IPv6 namespace, write -6 before the IP address or host name (-4 forces IPv4 instead), for example:

restrict -6 default kod limited nomodify notrap nopeer noquery
restrict -6 ::1    # ::1 is the IPv6 equivalent for 127.0.0.1

Lastly, specify the drift file (which keeps track of your clock's time deviation) and optionally the log file location:

driftfile /var/lib/ntp/ntp.drift
logfile /var/log/ntp.log

A very basic configuration file will look like this:

/etc/ntp.conf

server 0.arch.pool.ntp.org iburst
server 1.arch.pool.ntp.org iburst
server 2.arch.pool.ntp.org iburst
server 3.arch.pool.ntp.org iburst
tos orphan 15

restrict default kod limited nomodify notrap nopeer noquery
restrict -6 default kod limited nomodify notrap nopeer noquery

restrict 127.0.0.1
restrict -6 ::1  

driftfile /var/lib/ntp/ntp.drift
logfile /var/log/ntp.log

Note: Defining the log file is not mandatory, but it is always a good idea to have feedback for ntpd operations.

Usage

The package has a default client-mode configuration and its own user and group to drop root privileges after starting. If you start it from the console, you should always do so with the -u option:

# ntpd -u ntp:ntp

The -u option is employed by the two included systemd services. These services also use the -g option, which disables a threshold (so-called panic-gate). Hence, they will synchonize time even in case the ntp-server's time exceeds the threshold deviation from the system clock.

Warning: One reason the panic-gate was introduced is that background jobs/services may be susceptible to time-jumps. If the system's clock was never synchronized before, consider stopping them before running ntpd for the first time.

Both services are tied to the system's resolver, and will start synchronizing when an active network connection is detected.

Start ntpd at boot

Enable the daemon with ntpd.service. See also #Running in a chroot.

Note: The systemd command timedatectl can only be used to control systemd-timesyncd. Executing timedatectl set-ntp 1 as root will inadvertedly stop a running ntpd.service.[1]

Use ntpq to see the list of configured peers and status of synchronization:

$ ntpq -p

The delay, offset and jitter columns should be non-zero. The servers ntpd is synchronizing with are prefixed by an asterisk. It can take several minutes before ntpd selects a server to synchronize with; try checking after 17 minutes (1024 seconds).

Synchronize time once per boot

Alternatively, enable ntpdate.service to synchronize time once (option -q) and non-forking (option -n) per boot, instead of running the daemon in the background. This method is discouraged on servers, and in general on machines that run without rebooting for more than a few days.

If the synchronized time should be written to the hardware clock as well, configure the provided unit as described in systemd#Editing provided units before starting it:

/etc/systemd/system/ntpdate.service.d/hwclock.conf

[Service]
ExecStart=/usr/bin/hwclock -w

Tips and tricks

Start ntpd on network connection

ntpd can be started by your network manager, so that the daemon only runs when the computer is online.

Netctl

Append the following lines to your netctl profile:

ExecUpPost="systemctl start ntpd.service"
ExecDownPre="systemctl stop ntpd.service"

NetworkManager

The ntpd daemon can be brought up/down along with a network connection through the use of NetworkManager's dispatcher scripts. The networkmanager-dispatcher-ntpd^AUR package installs one, pre-configured to start and stop the ntpd service with a connection.

KDE

KDE can use NTP (ntp must be installed) by right clicking the clock and selecting Adjust date/time. However, this requires the ntp daemon to be disabled before configuring KDE to use NTP. [2]

Using ntpd with GPS

Most of the articles online about configuring ntpd to receive time from a GPS suggest to use the SHM (shared memory) method. However, at least since ntpd version 4.2.8, a much better method is available. It connects directly to gpsd, so gpsd needs to be installed.

Add these lines to your /etc/ntp.conf:

/etc/ntp.conf

#=========================================================
#  GPSD native ntpd driver
#=========================================================
# This driver exists from at least ntp version 4.2.8
# Details at
#   https://www.eecis.udel.edu/~mills/ntp/html/drivers/driver46.html
server 127.127.46.0 
fudge 127.127.46.0 time1 0.0 time2 0.0 refid GPS

This will work as long as you have gpsd working. It connects to gpsd via the local socket and queries the "gpsd_json" object that is returned.

To test the setup, first ensure that gpsd is working by running:

$ cgps -s

Then wait a few minutes and run ntpq -p. This will show if ntpd is talking to gpsd:

$ ntpq -p

remote           refid            st t when poll reach   delay   offset  jitter
 ==================================================================================
*GPSD_JSON(0)    .GPS.            0 l   55   64  377    0.000    2.556  14.109

Tip:

If the reach column is 0, it means ntpd has not been able to talk to gpsd. Wait a few minutes and try again. Sometimes it takes ntpd a while.
The GPS device must have PPS capabilities. You can test if your device is capable running ppscheck /dev/gps0.

Note: ntpd expects that your GPS device is called e.g. /dev/gps0. If your GPS device is connected via USB, it may appear as /dev/ttyUSB0 instead, and you may have to create a symlink ln -s /dev/ttyUSB0 /dev/gps0 and run gpsd on that linked /dev/gps0 so that the GPSD_JSON line appears as expected. gpsd should be run with the -n flag on the GPSD_OPTIONS line and use /dev/gps0 on the DEVICES line in the /etc/default/gpsd config file.

Running in a chroot

Note: ntpd should be started as non-root (default in the Arch Linux package) before attempting to jail it in a chroot, since chroots are relatively useless at securing processes running as root.

Create a new directory /etc/systemd/system/ntpd.service.d/ if it does not exist and a file named customexec.conf inside with the following content:

[Service]
ExecStart=
ExecStart=/usr/bin/ntpd -g -u ntp:ntp -i /var/lib/ntp

Then, edit /etc/ntp.conf to change the driftfile path such that it is relative to the chroot directory, rather than to the real system root. Change:

driftfile       /var/lib/ntp/ntp.drift

to

driftfile       /ntp.drift

Create a suitable chroot environment so that getaddrinfo() will work by creating pertinent directories and files (as root):

# mkdir /var/lib/ntp/etc /var/lib/ntp/lib /var/lib/ntp/proc
# mkdir /var/lib/ntp/usr /var/lib/ntp/usr/lib
# touch /var/lib/ntp/etc/resolv.conf /var/lib/ntp/etc/services

and by bind-mounting the aformentioned files:

/etc/fstab

...
#ntpd chroot mounts
/etc/resolv.conf  /var/lib/ntp/etc/resolv.conf none bind 0 0
/etc/services	  /var/lib/ntp/etc/services none bind 0 0
/lib		  /var/lib/ntp/lib none bind 0 0
/usr/lib	  /var/lib/ntp/usr/lib none bind 0 0
/proc		  /var/lib/ntp/proc none bind 0 0

# mount -a

Finally, restart ntpd daemon again. Once it restarted you can verify that the daemon process is chrooted by checking where /proc/{PID}/root symlinks to:

# ps -C ntpd | awk '{print $1}' | sed 1d | while read -r PID; do ls -l /proc/$PID/root; done

should now link to /var/lib/ntp instead of /.

It is relatively difficult to be sure that your driftfile configuration is actually working without waiting a while, as ntpd does not read or write it very often. If you get it wrong, it will log an error; if you get it right, it will update the timestamp. If you do not see any errors about it after a full day of running, and the timestamp is updated, you should be confident of success.

Restrict listening sockets

You can limit sockets ntpd is listening to using the interface option:

interface [listen | ignore | drop] [all | ipv4 | ipv6 | wildcard | name | address[/prefixlen]]

like

/etc/ntp.conf

interface listen lo
interface listen enp3s0
interface ignore enp5s0