Origin of Synchronisation (Part 2)

  |   By

Continued…

Most towns and cities would have a main clock, such as Big Ben in London, and for those living near-by, it was fairly easy to look out the window and adjust the office or factory clock to ensure synchronicity; however, for those not in view of these tower clocks, other systems were used.

Commonly, somebody with a pocket watch would set the time by the tower clock in the morning and then go around businesses and for a small fee, let people know exactly what the time was, thus enabling them to adjust the office or factory clock to suit.

When, however, the railways began, and timetables became important it was clear a more accurate method of time keeping was needed, and it was then that the first official time-scale was developed.

As clocks were still mechanical, and therefore inaccurate and prone to drift, society again turned to that more accurate chronometer, the sun.

It was decided that when the sun was directly above a certain location, that would signal noon on this new time-scale. The location: Greenwich, in London, and the time-scale, originally called railway time, eventually became Greenwich Meantime (GMT), a time-scale that was used until the 1970’s.

Now of course, with atomic clocks, time is based on an international time-scale UTC (Coordinated Universal Time) although its origins are still based on GMT and often UTC is still referred to as GMT.

Now with the advent of international trade and global computer networks, UTC is used as the basis of nearly all international time. Computer networks deploy NTP servers to ensure that the time on their networks are accurate, often to a thousandth of a second to UTC, which means all around the world computers are ticking with the same accurate time – whether it is in London, Paris, or New York, UTC is used to ensure that computers everywhere can accurately communicate with each other, preventing the errors that poor time synchronisation can cause.

Origin of Synchronisation (Part 1)

  |   By

Part One

With modern NTP servers (Network Time Protocol) synchronisation is made easy. By receiving a signals from GPS or radio signals such as MSF or WWVB, computer networks consisting of hundreds of machines can easily be synchronised together, ensuring trouble free networking and accurate time-stamping.

Modern NTP time servers are reliant on atomic clocks, accurate to billions of parts of a second, but atomic clocks have only been around for the last sixty years and synchronisation has not always been so easy.

In the early days of chronology, clocks mechanical in nature, were not very accurate at all. The first time-pieces could drift by up to an hour a day so the time could differ from town clock to town clock, and most people in the agricultural based society regarded them as a novelty, relying in stead on sunrise and sunset to plan their days.

However, following the industrial revolution, commerce became more important to society and civilisation, and with it, the need to know what the time was; people needed to know when to go to work, when to leave and with the advent of railways, accurate time became even more crucial.

In the early days if industry, workers were often woken for work by people paid to wake them up. Known as ‛knocker-uppers.’ Relying on the factory time-peice, they would go around town and tap on people’s windows, alerting them to the start of the day, and the factory hooters signalled the beginning and end of shifts.

However, as commerce developed time became even more crucial, but as it would take another century or so for more accurate timepieces to develop (until at least the invention of electronic clocks), other methods were developed.

To follow…

The Accuracy of the Speaking Clock

  |   By

The UK speaking clock has been around for nearly eighty years. It was started in 1936 when time keeping started to become more important to people’s lives. Initially available only in the London it was rolled out to the whole country during World War II.

There have been four people that have had honour of providing the permanent voice to the speaking clock over the last 70. And over 70 million calls are made to the speaking clock making it an important from of accurate time but have you ever wondered how accurate it is and where the time comes from and how accurate it is?

The speaking clock is controlled by a major British telecoms company who took over the General Post Office (GPO) and the time was originally supplied by the National Physical Laboratory (NPL) who also provide the MSF signal that NTP time servers use as a source of atomic clock synchronisation.

NPL no longer help with the speaking clock but the time is still controlled by NTP servers, either GPS or MSF, which ensures that the time you hear on the end of the telephone is accurate.

NTP servers are also commonly used by computer networks to ensure that IT systems, from traffic light signals to the office PC are all running an accurate form of time.

NTP time servers can either receive the MSF radio signal broadcast by NPL or, more commonly now, GPS signals beamed directly from space.

Often network administrators opt to use online NTP servers that send time signals over the internet but these are not as accurate and cause security problems so it is far better to have a dedicated NTP time server to control the time if you wish to have a computer network that is running accurately.

Finding an Online NTP Time Source

  |   By

Finding a source of time to synchronise a computer network to can be a challenge as there are a myriad of online time sources, all pertaining to be accurate and reliable; however, the truth can be rather different with many online sources either in too much demand, too far away or inaccurate.

NTP (Network Time Protocol) requires a source of UTC time (Coordinated Universal Time) which is kept true by atomic clocks. Online time sources are not themselves atomic clocks but NTP server devices that receive the time from an atomic clock which is then relayed to the devices that connect to the online time server.

There are two types of online time server: stratum 1 devices – devices that receive the time directly from an atomic clock, either using GPS or a radio reference signal. Stratum 2 devices  on the other hand are one step further away in that they are receive their time from a stratum 1 time server.

Because of demand, finding an online stratum 1 time server is next to impossible, and those that do take request usually do so under a subscription, which leaves the only choice for most people being a stratum 2 device.

There are plenty of resources on the internet that provide locations for online time servers.

But there are drawbacks to using such devices; firstly, online stratum 2 time sources can’t be guaranteed and several surveys taken have found that the reliability and accuracy of many of them can’t be taken for granted.  Secondly, online sources of time require an open firewall port which can be manipulated by malicious bots or users – leading to security risks.

A far better solution for most networks is to install your own stratum 1 NTP server. These time server devices sync to atomic clocks outside the firewall (using GPS or radio signals) and therefore are not security risks. They are also accurate to a few milliseconds ensuring the network will always be accurate to UTC.