Posted by Richard N Williams on February 2nd, 2009
NTP (Network Time Protocol) is an internet based protocol designed to synchronise the clocks on a computer network. It is the main time synchronisation software used in computer networks and is also packaged with most operating systems.
An NTP server is a dedicated device that receives a single time source then distributes it amongst all devices on a network. The protocol NTP monitors the drift of the internal clocks on a network and corrects for them.
An NTP server can receive a time source from either a national physical laboratory such as the UK’s National Physical Laboratory (NPL), however, these time signals are broadcast via long wave radio and have finite range.
GPS NTP servers are designed to receive the time source generated by the atomic clocks onboard GPS satellites (Global Positioning System). GPS is available anywhere on the planet as a time source as long as there is a clear view of the sky.
Without correct synchronisation all sorts of potential problems can occur such as leaving a computer system vulnerable to fraud, malicious users and hackers. An unsynchronised computer network may also lose data and be difficult to audit.
A global timescale called UTC (Coordinated Universal Time) has been developed to ensure the entire world uses the same timescale. The NTP server utilise UTC ensuring the computer network is telling the same time as every other computer network.
Posted by Stuart on January 30th, 2009
We may think of their being only one time and therefore one timescale. Sure, we’re all aware of time zones where the clock has to be pushed back an hour but we all obey the same time surely?
Well actually we don’t. There are numerous different timescales all developed for different reasons are too numerous to mention them all but it wasn’t until the nineteenth century that the idea of a single timescale, used y everybody came into effect.
It was the advent of the railway that provoked the first national timescale in the UK (Railway time) before then people would use noon as a basis for time and set their clocks to it. It rarely mattered if your watch was five minutes faster than your neighbours but the invention of the trains and the railway timetable soon changed all that.
The railway timetable was only useful if people all used the same time scale. A train leaving at 10.am would be missed if a watch was five minutes slow so synchronisation of time became a new obsession.
Following railway time a more global timescale was developed GMT (Greenwich Meantime) which was based on the Sun’s position at noon which fell over the Greenwich Meridian line (0 degrees longitude). It was decided during a world conference in 1884 that a single world meridian should replace the numerous one’s already in existence. London was perhaps the most successful city in the world so it was decided the best place for it.
GMT allowed the entire world to synchronise to the same time and while nations altered their clocks to adjust for time-zones their time was always based on GMT.
GMT proved a successful development and remained the world’s global timescale until the 1970’s. By then that atomic clock had been developed and it was discovered in the use of these devices that Earth’s rotation wasn’t a reliable measure to base our time on as it actually alters day by day (albeit by fractions of a second).
Because of this a new timescale was developed called UTC (Coordinated Universal Time). UTC is based on GMT but allows for the slowing of the Earth’s rotation by adding additional ‘Leap Seconds’ to ensure that Noon remains on the Greenwich Meridian.
UTC is now used all over the World and is essential for applications such as air traffic control, satellite navigation and the Internet. In fact computer networks across the globe are synchronised to UTC using NTP time servers (Network Time Protocol). UTC is governed by a constellation of atomic clocks controlled by national physics laboratories such as NIST (National Institute of Standards and Time) and the UK’s NPL.
Posted by Richard N Williams on January 28th, 2009
3. Security Breaches:
When networks are not synchronised log files are not recorded properly or in the right order which means that hackers and malicious users can breach security unnoticed. Many security software programs are also reliant on timestamps with anti-virus updates failing to happen or scheduled tasks falling behind. If your network controls time-sensitive transactions then this can even result in fraud if there is a lack of synchronisation.
4. Legal Vulnerability:
Time is not just used by computers to order events it is used in the legal world too. Contracts, receipts, proof-of-purchase are all reliant on time. If a network is not synchronised then it becomes difficult to prove when transactions actually took place and it will prove difficult to audit them. Furthermore, when it comes to serious matters such as fraud or other criminality a dedicated NTP server or other network time server device synchronised to UTC is legally auditable, its time can not be argued with!
5. Company Credibility:
Succumbing to any of these potential hazards can not just have devastating effects on your own business but also that of your clients and suppliers too. And the business grapevine being what it is any potential failing on your part will soon become common knowledge amongst your competitors, customers and suppliers and be seen as bad business practices.
Running a synchronised network adhering to UTC is not difficult. Many network administrators think that synchronisation just means an occasional time request to an online NTP time source; however, doing so will leave a system just as vulnerable to fraud and malicious users as having no synchronisation. This is because to use an Internet time source would require leaving a permanent port open in the firewall.
The solution is to use a dedicated NTP time server that receives a UTC time source from either a radio transmission (broadcast by national physics laboratories) or the GPS network (Global Positioning System). These are secure and can keep a network running to within a few milliseconds of UTC.
Posted by Richard N Williams on January 27th, 2009
Most businesses these days rely on a computer network. Computers in most organisations conduct thousands of tasks a second, from controlling production lines; ordering stock; preparing financial records and communicating with computers on other networks – often from the other side of the world.
Computers use just one thing to keep track of all these tasks: time. Timestamps are the computers only reference for when an event or task occurs in relation to other events. They receive time in the form of timestamps and they measure time in periods of milliseconds (thousandth of a second) as they may conduct hundreds of processes each second.
A global timescale known as UTC (Coordinated Universal Time) has been developed to ensure computers from different organisations all over the world can synchronise together. So what happens if the clocks on computers don’t coincide with each other or with UTC?
The consequences of running a network with computers that are not synchronised can be disastrous. Here are five reasons why all businesses need adequate network synchronisation using a NTP server (Network Time Protocol) or other network time server device.
1. Tasks fail to happen:
When computers are running at different times, events on different machines can fail to happen as often a PC may assume an event on another machines has already happened if the time for that event has passed according to its own clock. And what is worse, when one task fails it has a knock-on effect with other tasks failing to happen and in turn causing further tasks to fail.
2. Loss of Data:
When tasks fail to happen it soon gets noticed but when networks are not synchronised data that is meant to be kept can quite easily be lost and it can go unnoticed for quite a while. Data can be lost because storage as and retrieval is also reliant on time stamps.
Posted by Richard N Williams on January 25th, 2009
The atomic clock is the culmination of mankind’s obsession of telling accurate time. Before the atomic clock and the nanosecond accuracy they, employ time scales were based on the celestial bodies.
However, thanks to the development of the atomic clock it has now been realised that even the Earth in its rotation is not as accurate a measure of time as the atomic clock as it loses or gains a fraction of a second each day.
Because of the need to have a timescale based somewhat on the Earth’s rotation (astronomy and farming being two reasons) a timescale that is kept by atomic clocks but adjusted for any slowing (or acceleration) in the Earth’s spin. This timescale is known as UTC (Coordinated Universal Time) as employed across the globe ensuring commerce and trade utilise the same time.
Computer networks use network time servers to synchronise to UTC time. Many people refer to these time server devices as atomic clocks but that is inaccurate. Atomic clocks are extremely expensive and highly sensitive pieces of equipment and are only usually to be found in universities or national physics laboratories.
Fortunately national physics laboratories like NIST (National Institute for Standards and Time – USA) and NPL (National Physical Laboratory – UK) broadcast the time signal from their atomic clocks. Alternatively the GPS network is another good source of accurate time as each GPS satellite has onboard its own atomic clock.
The network time server receives the time from an atomic clock and distributes it using a protocol such as NTP (Network Time Protocol) ensuring the computer network is synchronised to the same time.
Because network time servers are controlled by atomic clocks they can keep incredibly accurate time; not losing a second in hundreds if not thousands of years. This ensures that the computer network is both secure and unsusceptible to timing errors as all machines will have the exact same time.
Posted by Richard N Williams on January 23rd, 2009
The atomic clock is the culmination of mankind’s ability to keep time that has spanned several millennia. Humans have always been preoccupied with keeping track of time ever since early man noticed the regularity of the celestial bodies.
The sun, moon, stars and planets soon became the basis for out timescales with periods of time such as years, months, days and hours based solely on the regulation of the Earth’s rotation.
This worked for thousands of years as a reliable guide to how much time has past but over the last few centuries humans have strode to find even more reliable methods for keeping track of time. Whilst the Sun and celestial bodies were an affective way sundials didn’t work on cloudy days and as the days and night s altered during the year only noon (when the sun is at its highest) could be reasonably relied upon.
The first foray into an accurate timepiece that was not reliant on celestial bodies and was not a simple time (such as a candle taper or water clock) but actually told time over a prolonged period was the mechanical clock.
These first devices dating as far back as the twelfth century were crude mechanisms using a verge and foliot escapement (a gear and lever) to control the ticks of the clock. After a few centuries and a myriad of designs the mechanical clock took its next step forward with the pendulum. The pendulum gave clocks their first true accuracy as it controlled with more precision the ticks of the clock.
However, it wasn’t until the twentieth century when clocks entered the electronic age did they become truly accurate. The digital and electronic clock had its ticks controlled by using the oscillation of a quartz crystal (its changed energy state when a current is based through) which proved so accurate that rarely a second a week was lost.
The development of atomic clocks in the 1950’s used the oscillation of a single atom which generates over 9 billion ticks a second and can maintain precise time for millions of years without losing a second. These clocks now form the basis of our timescales with the entire world synchronised to them using NTP servers, ensuring wholly accurate and reliable time.
Posted by Stuart on January 21st, 2009
Synchronising a network is often considered a headache by network administrators who fear that getting it wrong can lead to disastrous results and while there is no deny that a lack of synchronisation can cause unforeseen problems particularly with time sensitive transactions and security, perfect synchronisation is simple if these steps are followed:
1. Use a dedicated NTP server. The NTP server is a device that receives a single time source then distributes it amongst a network of computers using the protocol NTP (Network Time Protocol) one of the oldest Internet based protocols and by far the most widely used time synchronisation software. NTP is often packaged with modern operating systems such as Windows or Linux although there is no substitute for a dedicated NTP device.
2. Always use a UTC time source (Coordinated Universal Time). UTC is based on GMT (Greenwich Meantime) and International Atomic Time (TAI) and is highly accurate. UTC is used by computer networks all over the world ensuring that commerce and trade are all using the same timescale.
3. Use a secure an accurate time signal. Whilst time signals are available all over the Internet they are unpredictable in their accuracy and while some may offer decent enough precision an Internet time server is outside a networks firewall which if left open to receive a timecode will cause vulnerabilities in the security of the network. Either GPS (global positioning system) or a dedicated radio signal such as those transmitted by national physics laboratories (such as MSF – UK, WWVB – USA, DCF –Germany) offer secure and reliable methods of receiving a secure and accurate time signal.
4. Organise a network into stratum, levels. Strata ensure that the NTP server is not inundated with time requests and that the network bandwidth doesn’t become congested. A stratum tree is organised by a few select machines being stratum 2 devices in that they receive a time signal from the NTP server (stratum 1 device) these in turn distribute the time to other devices (stratum 3) and so on.
5. Ensure all machines are utilising UTC and the NTP server tree. A common error in time synchronisation is to not ensure all machines are properly synchronised, just one machine running inaccurate time can have unforeseen consequences.
Posted by Richard N Williams on January 19th, 2009
There are several timescale used throughout the world. Most NTP servers and other network time servers use UTC as a base source however, there are others:
When we are asked the time it is very unlikely we would respond with ‘for which timescale’ yet there are several timescales used all over the globe and each is based on different methods of keeping track of the time.
Greenwich Mean Time (GMT) is the local time on the Greenwich meridian based on the hypothetical mean sun. As the Earth’s orbit is elliptical and its axis is tilted, the actual position of the sun against the background of stars appears a little ahead or behind the expected position. The accumulated timing error varies through the year in a smoothly periodic manner by up to 14 minutes slow in February to 16 minutes fast in November. The use of a hypothetical mean sun removes this effect. Before 1925 astronomers and navigators measured GMT from noon to noon, starting the day 12 hours later than in civil usage which was also commonly referred to as GMT. To avoid confusion astronomers agreed in 1925 to change the reference point from noon to midnight, and a few years later adopted the term Universal Time (UT) for the “new” GMT. GMT remains the legal basis of the civil time for the UK.
Universal Time (UT) is mean solar time on the Greenwich meridian with 0 h UT at mean midnight, and since 1925 has replaced GMT for scientific purposes. By the mid-1950s astronomers had much evidence of fluctuations in the Earth’s rotation and decided to divide UT into three versions. Time derived directly from observations is called UT0, applying corrections for movements of the Earth’s axis, or polar motion, gives UT1, and removing periodic seasonal variations generates UT2. The differences between UT0 and UT1 are of the order of thousandths of a second. Today, only UT1 is still widely used as it provides a measure of the rotational orientation of the Earth in space..
The world time standard (UTC):
Although TAI provides a continuous, uniform, and precise time scale for scientific reference purposes, it is not convenient for everyday use because it is not in step with the Earth’s rate of rotation. A time scale that corresponds to the alternation of day and night is much more useful, and since 1972, all broadcast time services distribute time scales based on Coordinated Universal Time (UTC). UTC is an atomic time scale that is kept in agreement with Universal Time. Leap seconds are occasionally
Information courtesy of the National Physical Laboratory UK.
Posted by Stuart on January 18th, 2009
Apart from the usual celebrations and revelry the end of December brought with the addition of another Leap Second to UTC time (Coordinated Universal Time).
UTC is the global timescale used by computer networks across the world ensuring that everybody is keeping the same time. Leap Seconds are added to UTC by the International Earth Rotation Service (IERS) in response to the slowing of the Earth’s rotation due to tidal forces and other anomalies. Failure to insert a leap second would mean that UTC would drift away from GMT (Greenwich Meantime) – often referred to as UT1. GMT is based on the position of the celestial bodies so at midday the sun is at its highest above the Greenwich Meridian.
If UTC and GMT were to drift apart it would make life difficult for people like astronomers and farmers and eventually night and day would drift (albeit in a thousand years or so).
Normally leap seconds are added to the very last minute of December 31 but occasionally if more than one is required in a year then is added in the summer.
Leap seconds, however, are controversial and can also cause problems if equipment isn’t designed with leap seconds in mind. For instance, the most recent leap second was added on 31 December and it caused database giant Oracle’s Cluster Ready Service to fail. It resulted in the system automatically rebooting itself on New Year.
Leap Seconds can also cause problems if networks are synchronised using Internet time sources or devices that require manual intervention. Fortunately most dedicated NTP servers are designed with Leap Seconds in mind. These devices require no intervention and will automatically adjust the entire network to the correct time when there is a Leap Second.
A dedicated NTP server is not only self-adjusting requiring no manual intervention but also they are highly accurate being stratum 1 servers (most Internet time sources are stratum 2 devices in other words devices that receive time signals from stratum 1 devices then reissue it) but they are also highly secure being external devices not required to be behind the firewall.
Posted by Stuart on January 16th, 2009
Time synchronisation is often described as a ‘headache’ by network administrators. Keeping computers on a network all running the same time is increasingly important in modern network communications particularly if a network has to communicate with another network running independently.
For this reason UTC (Coordinated Universal Time) has been developed to ensure all networks are running the same accurate timescale. UTC is based on the time told by atomic clocks so it is highly precise, never losing even a second. Network time synchronisation is however, relatively straight forward thanks to the protocol NTP (Network Time Protocol).
UTC time sources are widely available with over a thousand online stratum 1 servers available on the Internet. The stratum level describes how far away a time server is to an atomic clock (an atomic clock that generates UTC is known as a stratum 0 device). Most time servers available on the Internet are in fact not stratum 1 devices but stratum in that they get their time from a device that in turn receives the UTC time signal.
For many applications this can be accurate enough but as these timing sources are on the Internet there is very little you can do to ensure both their accuracy and their precision. In fact even if an Internet source is highly accurate the distance away form it can cause delays int eh time signal.
Internet time sources are also unsecure as they are situated outside of the firewall forcing the network to be left open for the time requests. For this reason network administrators serious about time synchronisation opt to use their own external stratum 1 server.
These devices, often called a NTP server, receive a UTC time source from a trusted and secure source such as a GPS satellite then distribute it amongst the network. The NTP server is far more secure than an Internet based time source and are relatively inexpensive and highly accurate.