I use Network File System (NFS) on my Linux systems to share some disk space among many machines on my network. NFS is a distributed file system protocol originally developed by Sun Microsystems in 1984, allowing a user on a client computer to access files over a network much like local storage is accessed.

NFS doesn't synchronize time between client and server, and offers no mechanism for the client to determine what time the server thinks it is. What this means is that a client can update a file, and have the timestamp on the file be either some time long in the past, or even in the future, from its point of view.

While this is generally not an issue, if clocks are a few seconds or even a few minutes off, it can be confusing and misleading to humans. Of even greater importance is the effect this has on programs. Programs often do not expect time differences like this, and may end abnormally or behave strangely, as various tasks timeout instantly, or take an extraordinarily long while to timeout.

Poor time synchronization also makes debugging problems difficult, because there is no easy way to establish a chronology of events. This is especially problematic when investigating security issues, such as break in attempts.

The workaround is to use Network Time Protocol (NTP) religiously. Use of NTP can result in machines that have extremely small time differences. Note: The NFS protocol version 3 does have support for the client specifying the time when updating a file, but this is not widely implemented. Additionally, it does not help in the case where two clients are accessing the same file from machines with drifting clocks.

Installation

Rather than use the ntpd or the Network Time Protocol daemon, I opted to use a much simpler and easier to use implemention: Chrony. Chrony implements the NTP protocol and can act as either a client or a server. I use them as both. Here's how I'm doing it...

First, install chrony ... In Debian, simply run the following command

1

$ sudo apt-get install chrony

There are 2 main programs that come with chrony: the client and the server or chronyc and chronyd respectively. chronyd is a daemon which runs in background on the system. It obtains measurements via the network of the system clock’s offset relative to time servers on other systems and adjusts the system time accordingly. chronyc provides a user interface to chronyd for monitoring its performance and configuring various settings. It can do so while running on the same computer as the chronyd instance it is controlling or a different computer.

Configuration

All of my systems use both chronyc and chronyd. I have one server that looks to the Internet for its time source and all other machines point to this one server for their sources. You can achive this setup via configuration. The first thing you need to setup is the external servers you are going to point to. Here's what I am using:

1
2
3
4
5
6
7
8

server 0.arch.pool.ntp.org iburst
server 1.arch.pool.ntp.org iburst
server 2.arch.pool.ntp.org iburst

server 0.debian.pool.ntp.org iburst minpoll 8
server 1.debian.pool.ntp.org iburst minpoll 8
server 2.debian.pool.ntp.org iburst minpoll 8
server 3.debian.pool.ntp.org iburst minpoll 8

And then all other servers on my network use the following:

1

server 10.0.0.92

The above IP is my Chrony time server. Here are some of the other most interesting configuration items for Chrony:

1. The driftfile stores the computer's clock gain/loss rate in parts per million.

   1	driftfile /var/lib/chrony/drift

2. The program called chronyc can configure aspects of chronyd operation once it is running. You can protect it with a password stored in the keys file. You also need to tell chronyd which numbered key in the file is used as the password for chronyc.

   1 2	keyfile /etc/chrony.keys commandkey 2

3. If you want to act as an NTP server for other computers. You might be running chronyd on a machine that has a LAN sitting behind it with several computers on it. By default, chronyd does not allow any clients to access it. You need to explicitly enable access using allow and deny directives.

   1	allow 10.0.0.0/24

4. You may also want chronyd to act as a NTP broacast server. This means that a broadcast packet is sent to the broadcast address every 60 seconds. The ip address MUST correspond to the broadcast address of one of the network interfaces on your machine.

   1	broadcast 60 10.0.0.255

5. If you want to present your computer's time for others to synchronise with, even if you don't seem to be synchronised to any NTP servers yourself, enable the following line in the config file. The value 10 may be varied between 1 and 15. You should avoid small values because you will look like a real NTP server. The value 10 means that you appear to be 10 NTP 'hops' away from an authoritative source (atomic clock, GPS receiver, radio clock etc).

   1	local stratum 10

6. Beyond the above settings, you can also set the values for log and PID files, allow remote management, etc.

The Client

You may also want to interract with your chronyd server in real-time, and you can do that with the chronyc program. By calling simply chronyc on the command line you will connect to the local host. If you want to connect to a remote chronyd host, runt he following command:

1

$ chronyc -h 10.0.0.92

The above connects you to the chronyd server and presents a prompt where you can enter commands. Here are some of the most common sommands and their meanings:

1. password: Provide password needed for most commands
2. activity: Check how many NTP servers/peers are online/offline
3. clients: Show clients that have accessed the server
4. makestep: Immediately correct the system clock instead of slewing
5. online: Warn that connectivity to a source has been restored
6. sources: Display information about the current set of sources
7. sourcestats: Display the rate & offset estimation performance of sources
8. tracking: Display system clock performance

For further informatin and more detailed documentation go to Chrony home.