For non-techies facing business decisions
Now's the time to
Picture this: Jane and Judy have both worked hard to be successful. Each has built her own small but growing professional services firm.
Like most business owners, they rely every work day on a computer network. It's how the team connects to the internet, sends emails, and works on their projects.
For Jane, it's all smooth sailing. Things just work and honestly, she doesn't really have to think about it.
But for Judy? Constant crisis. Devices drop out, files don't save, it gets real slow, where'd those bloody printers go?
Sometimes, thanks to hacking or ransomware, her whole office shuts down.
How did they end up in such different places?
Ever been in the market for a new house or car?
Most of us are too busy with our real jobs to build one from scratch.
But still, you'd probably want some idea of what you're getting for your money - what you can do with it, what it's like to drive, and whether it's the right size for your family.
That's where this guide fits in. It's for all the lawyers, travel agents, accountants, hoteliers, and anyone else who has no desire to become an IT expert, and yet still relies on a working computer network to keep their business running.
It's especially meant for business owners who've outgrown the home office, but aren't yet large enough to warrant an in-house IT department. It's about making better decisions and knowing how to get the right help.
Because it's so not worth the aggravation and headache of getting it wrong.
You might not even realise that you have already built a computer network.
You know that wireless router that connects your laptop, phone and PlayStation to the internet? That's actually a small network in its own right.
Technically, it's known as a peer-to-peer network - rather than connect to a central server, all the computers share resources directly with each other.
So what's stopping you from using one for your business?
For the very smallest businesses, it can be a viable option. It's not the most sophisticated network, but it might be all you really need.
So what does such a simple network have going for it?
Let's jump right in.
Wireless routers start at under $100; sometimes your internet provider just gives you one. There's no need to hire a contractor to install wiring.
It doesn't take sophisticated IT knowledge to configure these things - if you do need to hire someone, the most junior technician can handle it.
All it really takes to get one of these things going is a trip to the office supplies shop and a basic willingness to read the instructions.
There's only one box where things can go wrong, and most problems can be fixed by switching it off and then on again.
If everyone could make do with the very cheapest and simplest kind of network, there'd be no reason to opt for anything else.
Here are the problems you can run into with this kind of network:
In a peer-to-peer network, every user is more or less equal. You can't, for instance, give your bookkeeper and your operations manager access to the financials without also giving the rest of the office the same access.
In a really small business, this might not be too bad. It might be practical to keep all your financials on one computer and not share them on the network. If you're using cloud services for your accounts, this might not matter at all.
These kind of networks also tend to suffer from the ad hoc nature of the IT support they receive. You might rely on someone in the office who has a different job but who “knows computers”, or you might wait until you notice something goes wrong before you call someone in. Either way, they won't be proactive about keeping everything patched and updated for security.
In this kind of network, every computer loads and saves work to a local hard drive, just the same as if it wasn't connected to a network at all.
You might end up collaborating on projects by emailing each files to each other. This can lead to version control problems where everyone has their own version of the same file, with different edits made at different times.
This can also make backups more complicated. Instead of backing up one file server, every hard drive needs to be backed up.
One way you can address both problems is by using a cloud service such as Dropbox, which synchronises project directories across all users and to the cloud.
Another way to address this is to centralise your files to a network attached storage device, but this will make your network more complicated to set up, and it will cost more.
The theoretical limit of most wireless routers is between 30 and 255 devices.
As a practical matter, speed and performance will degrade long before you hit this mark. It will get slow and devices will drop out.
And because every computer on the network has to act as a server to every other computer, they all have to work harder as the network grows larger.
You generally don't want to stretch one of these networks to more than about 12 devices. That includes phones, tablets and printers.
Your operating system also has limits on how many users can connect to the same peer-to-peer network - most newer versions of Windows are limited to 10.
This is a good, cheap way for the very smallest businesses to share printers, an internet connection and send files to each other.
And if that's all you need for the moment, then there might not be much reason to consider anything else.
For everyone else, it's time to consider a client/server network (more on that in the next section).
In a client/server network, you have a single central computer - known as the “server” - managing all the other computers and mobile devices in your office - the “clients”.
There are a number of advantages to this approach:
With a client/server network, everybody can save their work onto the server. This eliminates the problem of different versions of the same file floating around as people email their work to each other.
It also makes backups way easier - and it makes it easier to keep both onsite and offsite backups.
This also means you have far less to lose if any of your desktop or laptop computers fail. If the whole machine goes up in smoke - just get a new one, install the software you need, and you're up and running again.
Storing all your work on the server also means you can have 24x7 access to your files.
With a client/server network, you can give different users different levels of access to different files, databases and other services.
That means you can set it up so that only you, your operations manager and your accounts guy are able to access the financials. If you handle commercial-in-confidence information, then you can also set it up so that only people working on a particular project have access to it.
And if you provide free wi-fi to the public, then you don't have to let them have access to all your credit cards on file.
In a peer-to-peer network, every computer acts as a server to every other computer on the network. As the network grows larger, this means more work for every computer.
In a client/server network, the server takes all this load, leaving the workstations to just run the programs your staff use to deliver their work.
This is the big one: where peer-to-peer networks start to choke as they hit a certain size, client/server networks can be scaled.
There are two kinds of client/server networks: “thin client/fat server” networks and “fat client/thin server” networks.
A fat client is the kind of computer you're probably most familiar with - one like you'd have at home. It can run programs on its own hardware.
A thin client, on the other hand, is a stripped-down, bare bones computer whose only job is to connect to a server that runs all the applications.
The advantage of thin client networking is that thin clients require little to no maintenance. Technicians can manage and update everything on the server.
In recent years, the explosion in cloud services has meant that a lot of business software runs neither on the client nor the local server, but somewhere on the internet. In theory, if you can run everything on the cloud, all you need your client or server to do is keep you connected.
But as a practical matter, most small businesses end up wanting to run a mix of desktop applications, server applications and cloud-based services.
And while it's technically possible to run all the same desktop applications on a thin client, it's not quite as simple as just downloading the software and installing it.
This can be a pain for a lot of small business owners, who might expect to be able to easily try out new software packages from time to time.
Fat client networks are generally a good default option for most small businesses.
Thin clients will make sense for special use cases.
Back in the medieval period of computer networking, you'd need a separate computer for your file server, mail server, print server and so on.
These days, servers are easily powerful enough that all of these separate roles can be performed by virtual machines running on the one physical server.
For ordinary business needs - such as sharing and managing files, sharing printers, connecting to the internet and so on - one physical server will be enough for an office of hundreds of people.
One way to ensure uptime is to have a second, redundant server in your rack that can take over if the hardware fails.
This is an expensive option. And it won't help with many of the things can interfere with the proper use of your server, such as power outages, malware, floods and fires.
It's often more affordable to consider other measures to ensure uptime, such as server monitoring, regularly scheduled maintenance, a realistic retirement date for the hardware, and a way to access your data in the cloud in the event of a disaster.
But if every minute of downtime hurts you enough to justify the expense, you may also want to consider a failover server.
There's two schools of thought on how to look after your servers.
One is to keep it running until you notice something goes wrong.
This is not such a bad strategy - so long as someone is responsible for checking on it periodically.
Let's say one of the hard drives breaks and your server is now relying on the spare one. If someone notices this, it's easy enough to just replace the broken hard drive - it's a very minor hassle - nobody's work needs to be disturbed.
On the other hand, if nobody's checking this stuff, nobody will realise anything's wrong until the second drive breaks and all of a sudden you don't really have a file server anymore. Then nobody will be able to access their work until the thing gets fixed and everything has been restored from backups.
Your IT technician should be able to keep an eye on this kind of stuff remotely.
The second school of thought is that servers should receive regular maintenance. If every hour of downtime is expensive for you, then this is a good idea.
The two pieces of your server most likely to wear out are the fan and the hard drive.
Because they use moving parts and are in constant use, it's not a matter of whether they'll wear out, but when.
The hard drive holds all your data, so there are some obvious problems with this breaking. Yes, you can restore from your backups. But that takes time.
Because this is such a pain point, nearly every server uses a technology called RAID to write data to 2 or more hard drives so that if one drive fails, there's another to fall back on.
These are often hot swappable - this means that the failed drive can be removed and replaced without needing to switch the server off or interrupt anyone's work.
It's a good idea to attend to these failed hard drives ASAP. The right RAID setup is a good way to stop hard drive failures from interrupting everyone's work day, but this only goes so far. The remaining hard drive will be just as old and have gone through the exact same use as the one that just failed.
Just about every server and every networking device is enabled to use the Simple Network Management Protocol, or SNMP.
This can be configured so that your technician can be immediately notified if anything breaks or goes wrong.
If you have nowhere else to put it, your server and other networking devices can just end up on a desk or shelf by default.
This is easy and costs nothing to set up.
But what's likely to happen over time is that your server and other devices become a big messy pile of boxes and cables.
Navigating this rat's nest then becomes a huge waste of the technician's time.
In the long run, it tends to be a lot cheaper to mount the devices in a rack. That way, everything can be neatly organised.
It used to be that if you wanted to network your office, you had to connect everything via cable. The wireless technology that first arrived in the late 1990s was ultra slow, had a pathetic range, and drove you nuts with all the dropouts.
These days, wireless networks are ubiquitous in homes, offices and even public spaces.
There's a lot to be said for wireless network. You avoid most of the expense of installing wiring, as well as most of the hassle and disruption to your office. (You will probably still need to run some wiring between your wireless access points and the devices at the heart of your network.)
So far so good, right?
And given that many of our most frequently used devices - such as phones and tablets - don't even have a socket for an ethernet cable, why use anything but wireless?
It's actually not an unreasonable option for some businesses. But it becomes less and less feasible as your business grows.
Some of the newer wireless access points have a theoretical maximum speed of over 3000 megabits per second.
But you'll never actually get anywhere near this kind of speed in the real world. 400 megabits per second is more realistic.
So why do they advertise a speed you'll never have? Is it just some sleazy marketing gimmick?
It's more that wireless speeds vary so wildly in different conditions that the theoretical maximum speed is the only way to have any baseline to compare the performance of different devices.
An ordinary ethernet cable has a maximum speed of 1000 megabits per second. And unlike with the wireless equipment, this is a maximum speed you can actually enjoy.
It's not just that the wireless network is slower. It's that this speed is shared between all devices.
When you connect multiple devices to a wireless network, it seems like everything communicates simultaneously. But, at the micro level, what actually happens is that from moment to moment, the wireless access point is talking to only one device on the network at any one time. It juggles between them so fast that it seems like it's all happening at the same time.
Add too many devices, it becomes like a busy airport with one runway.
If your wireless network is just for mobile devices, there'll be plenty of bandwidth to go around.
Streaming an HD video to your phone or tablet consumes about 5 megabits per second. There's enough bandwidth on a wireless network for more than a few devices to stream 5 megabits per second.
There's enough bandwidth there for a few desktop and laptop computers as well.
But these computers can put a lot more traffic through the network. You'll be loading large project files and email attachments, and synchronising your work with your file server or to the cloud.
As you add more and more of them, things will start to lag and slow down.
Most wireless access points have a theoretical limit of 250 devices. That's enough for just about anyone.
But you're going to hit problems well before 250.
For a start, there's radio interference: more devices sending radio signals mean more of it, leading to lower speeds and dropouts.
There's also the fact that networks don't just carry the big bits of data that you see on your computer and phone.
For the network to work, there are lots of small messages flying around just to organise everything. There are messages to say that you're about to send something and how big it will be, and replies to these messages to say okay. There are messages sent to check that none of the data got corrupted while it was sent.
And the whole thing has to be encrypted so that your nextdoor neighbour can't listen in and read your emails.
There's even a bit of “Are you still here?” “Yes, I'm still here” going back and forth just to let everyone know you're not disconnected.
Add too many devices and this all just gets too crowded.
If you're running a bar or coffee shop, you might find your complimentary Wi-Fi slows to a crawl during the busiest periods.
Your customers might appreciate that you've at least made the effort, but if it's too slow to stream their cat videos, then what's the point?
Here's how to get things back on track:
Consumer grade wireless access points are built to be used at home. And for 3 or 4 people, they've got plenty of grunt for everyone.
But they'll struggle to carry a crowd.
You'll do a lot better with an access point built for enterprise use.
It really does depend a bit what you're doing to know how.
But as a rough rule of thumb, we find a good enterprise grade access point can handle 25 devices, while a consumer grade one starts to stuggle with 10.
Adding more access points means each one has fewer devices to handle.
It also means less distance between the device and the access point, increasing the speed of the connection.
At a certain size, you're just going to have to bite the bullet.
You should install wiring for your desktop PCs, printers and other large devices that aren't going to move around a lot.
That saves your wireless bandwidth for the portable devices that need it.
There are two kinds of cabling in standard use across small business networks: Category 5e and Category 6, more commonly called “Cat5e” and “Cat6”.
Cat5e is rated for 1000 megabits per second of data up to about 100 metres. Cat6 can carry a whopping 10,000 megabits per second, but only across a shorter distance.
Many computer networks will use a mix of both: the Cat5 cable is plenty fast enough for printers and workstations, while Cat6 can make a difference to the highest traffic connections at the heart of the network.
Most ethernet cables are intended for indoor use only.
They aren't built for rain and sunlight. When installers cut corners by using indoor cables outside, they tend to wear out.
If you need to connect two buildings together, or to run a length of cable along the outside of the building to reach a certain room, you should be sure to use outdoor cables.
If your server rack has a big ball of cables coming out of it, veering off in every direction, like spaghetti and meatballs, then anyone who has to work on it is going to be wasting a lot of time dealing with that.
To be honest, whenever we inherit this mess, the way we usually deal with it is to just unplug everything and put it back together properly. That's the quickest way to sort out what's going on.
Messy bunches of cables also indicate that the other work is likely to be sloppy or haphazard.
There should be some record of how everything was connected. If new technicians have to work this out as they go, you're wasting a lot of time.
The unfortunate reality is that attempts to hack your network are not a matter of if but when.
Why would someone you've never met try to hack your network?
For these reasons and more, hacking attempts are unavoidable.
So what should you do?
If you depend on a computer network for your livelihood, you want it overseen by professional network engineers.
Where so many small businesses get this wrong is by relying on informal IT support from someone in the office who has a different job but “knows computers”
Even if they really are good with computers, they're busy with their real job and will only attend to problems after they're discovered - ie, after you've been hacked.
And because they're not dealing with IT support most of the time, they're not “in the loop” with all the news about newly discovered threats and vulnerabilities.
The other way small businesses get this wrong is by hiring a “professional” who is still learning on the job. Look for a team led by someone with many years of experience building and running networks.
And make sure that it's part of your arrangement to keep your network's security up to date.
If you take a “we'll call you when we need you” approach with your technicians then network security will only be dealt with after you've noticed that there's been a problem.
Hackers also gain access by tricking you and your staff into clicking malicious links or email attachments. Using techniques like spear phishing attacks they can trick anyone in the office into handing over passwords or running malicious scripts.
For this reason, you should be sure that your staff are security smart about malware, suspicious emails and attachments.
Don't just assume that this stuff is obvious to everyone. It might be obvious to you - but different people have different backgrounds and levels of interest in computers and the internet. Hackers only keep trying these tricks because some of the time they work.
“Firewall” - it kinda sounds like a spell a wizard would cast in Game of Thrones.
It's actually a piece of software or hardware that prevents unauthorised access to your network.
For your firewall to be of any use, it needs to be configured. It might make you feel better just to have one in your server room. But until your technician configures it, it's a bit like having the world's most uncrackable safe and then leaving the door open.
Your software that's installed on a small chip in your device, controlling its basic functions.
From time to time, the manufacturer will release updates to patch any security vulnerabilities that are discovered.
This is a crucial task that tends to get skipped when you rely on informal IT support.
Make sure that sensitive parts of your network can only be accessed by those who really need it.
Your financials, for instance, should be only accessible to those who need it - you, your bookkeeper, your operations manager.
This helps limit the damage a disgruntled employee or former employee can do.
How easily can your passwords be guessed?
A lot of small businesses have no real password policy at all. It's kinda just “choose whatever you feel like on your first day and keep using it forever”.
This makes it almost certain that at least one member of staff has picked a password that can be cracked with a simple Dictionary attack.
So insist on strong passwords and change them regularly.
Well encrypted data means that even if the bad guys manage to eavesdrop on your network traffic or access your files, they won't be able to make sense of any of it.
You can encrypt your file server and use an encrypted connection for email.
The best firewall and strongest passwords will be no help at all if any member of the public can wander up to your workstation with a USB drive full of Ransomware during your lunch break.
Wireless networking adds to your security concerns. Get it too wrong, and just anyone can park outside your building and access all your stuff without even leaving the car.
Too many wireless networks have the default name the device came with. This means bad guys can find out at a glance what equipment you're using and all the vulnerabilities that have been discovered for that device.
Don't rename it to your business name! This also makes it too easy for someone who wants to harm you.
A good wireless passkey should have at least 20 characters long, including numbers, letters and symbols.
Wireless devices broadcast their signals into the surrounding atmosphere. This means that you can't really stop anyone nearby from listening in.
This is why it's so crucial for this data to be properly encrypted.
Use a modern encryption method, like WPA2 AES.
If you're shutting the office down while everyone's on holiday, turn the wireless network off.
We all know it's true: time is money, and downtime angries up the blood.
Which is why the price tag on your new computer network is really only part of the cost.
What's at least as important is how many working hours it delivers.
For a rough, back-of-the-envelope idea of how much downtime will really cost you, have a think about how much you spend per hour on rent and wages.
It's a little harder to measure what it does to morale and momentum. But this is still also a cost. As is the hit you take to your reputation every time you have to tell your customers they have to wait because the network's down again.
The exact cost of every hour of downtime will vary from business to business. The more each hour costs you, the greater the investment warranted in a computer network that stays up and that can recover fast from any disasters that hit your office.
Here's the deal: there's two kinds of IT professional.
There are the technology purists, and the results-driven pragmatists.
The purists tend to be excellent with computer tasks - but the business impact of their work doesn't really matter to them. When the network is built, they'll look over at a well configured server and firewall and smile for a job well done. The amount of project hours you've been able to bill to your customers while it was built is someone else's worry.
Purists aren't bad people - different people have different interests and it'd be a dull world if they didn't.
It's just that you really don't want them in charge of your network.
The pragmatists understand that while IT can be fascinating in its own right, in business it's the outcomes that count.
You can sniff these guys out - they will ask you about your business objectives, and how they can do the work with the least disruption to your business - you won't need to be prompt them to raise these issues.
If you don't make other plans, it's easy to fall in to a de facto project manager role. You might have a relationship with a technician, but they won't do anything until you call them in.
Because you're busy running the rest of your business, you'll tend to only deal with the network when a problem arises. When this happens there's no real strategy or planning guiding these decisions - it's all just about responding to crises and putting out fires.
It's much better to have a clear plan - who's in charge of keeping it secure, when the servers are going to get their scheduled maintenance, how long you will be using it before it all gets replaced, that sort of thing.
That way, you turn downtime into a thing that you manage, rather than a thing that the universe throws at you at the worst possible time.
One sure way to lose precious hours of productivity is to phone up an IT support provider and ask something like “Can you come in next week to update the server?”
(What does “update the server” even mean?)
The risk here is that the technicians will miss some small but crucial detail they weren't told about. This will mean you'll need to get them back in, and in the meantime your work is being disrupted.
You can avoid this by properly documenting and planning your moves to new infrastructure.
Ideally, this would mean planning not only for how the final configuration is to look, but ways to keep your staff productive while the work is being done and in the case of any unexpected hiccups.
Such measures could include moving data to the cloud while a new file server is being configured, so that your team still has access to their projects during the move.
One of the most straightforward ways to minimise disruption is to schedule installations, upgrades and maintenance work for times when there's less work to disrupt.
Here in Melbourne, a lot of businesses tend to slow down after Christmas, when staff and customers are on their summer holidays. But if you're running a beach resort, winter might be your quiet time.
Whenever these quiet times are, they're the ideal time to knock over IT projects. That way, everything is sorted for the busy period.
Yes, it's usually a higher hourly rate to get IT technicians and wiring contractors to work outside office hours.
You need to balance this extra expense against the real cost of getting the work done during office hours.
Networks can be set up so that technicians can log in and administer them remotely.
This is a good thing.
It means technicians need a lot fewer trips to the office - which, one way or another, you'll be paying for.
And it means a faster response time to any problems that arise.
A sudden loss of power can play havoc with your server. You can lose files. You can mess up the operating system.
If it happens right in the middle of writing to a database, the database can be corrupted. There's even a risk of expensive hardware damage.
These are all terrible headaches that require time and money to fix.
An Uninterruptible Power Supply, or UPS, is at heart just a big battery that sits between your server and the mains power.
As well as providing backup power, most units also provides a level of surge protection and voltage regulation, keeping your server safe from any nasty spikes in the power.
One thing business owners commonly misunderstand about a UPS is that they think it will keep the whole office running during a blackout. While this is technically possible, it's not how they're commonly used.
Usually, the UPS is only connected to the server and to other equipment needed to back up and shut down safely. Because batteries eventually run out, if you can't afford a single minute of downtime when the power is out, you may want to consider a generator backup.
The more common use of a UPS is to keep the server up and running during those momentary power outages that might only last a few seconds, and to give the server the chance to back up and shut down safely during longer power outages.
This means that power outages will still interrupt some work, but they won't harm your server. This means you can bounce back with no lasting problems when the power is restored.
What would you do if hackers took control of all your systems? Or if lightning struck the server room?
Would you be frozen by anxiety? Or would you know exactly what to do next?
Here's where it makes all the difference to have a Disaster Recovery Plan - a written document listing all the steps to take.
This saves your technician so much work in identiying what needs to be done - right at a time where every ticking second counts.
It also saves you the need to make important decisions in a moment of pressure, panic and confusion.
This process also helps you identify any additional preparations you could take.
Most business owners appreciate the need for backups. What's less well understood is that there's more to backups than just whether or not you have them.
It's not just about being able to recover your data: it's about how fast you get back on your feet.
The longer the recovery process takes, the more of the technician's time you'll have to pay for, and the more you'll spend in rent, wages and bills during your lost productivity.
Certain backup methods, such as Microsoft's Windows Server Backup Tool, are notorious for reporting that backups have been succesfully taken, only to turn out to be corrupted when it's time to restore.
Usually, this is not a complete show stopper. Your technician will be able to piece it all together eventually. But instead of taking hours, it might take days.
Depending on the internet speed you enjoy, and the amount of data you have to restore, it might get very slow having to push all of it through your internet connection.
It's kind of like filling a swimming pool through a drinking straw: it will fill, but it might test your patience.
Here's where it makes so much sense to have an onsite backup to restore from.
But it's not a great idea to rely solely on onsite backups. Otherwise you could lose all your data in a lightning strike, theft or fire.
In certain disaster situations - say, if your server room was flooded - it might be a while before things are ready to even start the data recovery process.
So what if you've got crucial documents you need ASAP?
Here's where it can really help to have a backup system in place that will let you access your files in the cloud.
It used to be that your computer network was just for your computers, and maybe a printer or two. The explosive growth of network-connected devices - phones, tablets, security cameras, even coffee machines - means that IP addresses should be managed for growth. Otherwise you will run out of addresses for everything.
It'd be nice to just “one and done” all this computer network stuff.
But just like the phone in your pocket, the day will come when it's time is over.
Advances in technology mean your requirements will change as newer and better software puts more demands on your hardware.
The hardware companies realise this - that the day will come when it will all be obsolete. So they engineer their products to last years, but not forever. Over time, heat and electricity of daily use will wear the equipment out.
The short answer is 3-5 years.
It's a similar lifespan to most desktop and laptop computers.
It's helpful to keep this in mind when thinking about how far ahead to plan for. You might have goals to one day have hundreds of employees. But if the business is you and your brother who have just moved out of the home office and hired your first employee, perhaps you don't need to design a network for that much growth just yet.
It's also helpful to keep this in mind when considering how to respond when something breaks. If you're 6 months away from replacing everything, then it's probably a good idea to just go with the quickest and cheapest thing to keep your network going. On the other hand, if the entire network has reached a certain age, it might be time to replace the whole thing.
The best way to avoid the hassles and headaches of an aging network is to mark a date in the calendar for the whole network to be replaced.
This means that instead of every single device choosing its own moment to disrupt everyone's work when it eventually breaks, you can replace all of it at a time you can schedule and plan for.
This also means that you get full advantage of improved technology: every device can connect to every other device using the most recent standards.
That's much better bang for your buck than paying for brand new shiny hardware you can't use to its full potential.
Of course, there's nothing all that magical about the stroke of midnight on your network's coming of age. When you show up to work the next day, it will all still be there, ticking over exactly as it did the day before.
So why not just keep it running?
As well as saving the expense of buying and installing new systems, there's also the advantage that the thing is already there and working: you don't have to do anything.
The obvious problem is that things are going to start breaking. You don't get to choose exactly when - it could happen hours away from a crucial deadline.
And because all the bits and pieces are just as old as each other - and engineered for the same kind of lifespan - it's likely that a lot of it might start breaking all at a similar time.
Inevitably, some of them will break at the worst possible time. This will put you on the hook for emergency call-out fees and cost you lost hours of productivity right when you need them most.
In the end, this time and money saving strategy saves you neither.
If you keep patching your, you'll end up with what we've nicknamed a Frankenstein Network - it's sewn together from various bits Frankenstein's monster.
If you connect blazing fast new technology to ancient relics, your network can only run as fast as the old stuff. So you're spending good many for the latest technology, without being able to enjoy it.
These networks tend grow more and more complicated as they accumulate layers of quick fixes that were meant to last a week or two but ended up hanging around for years. This makes them hard to grow and confusing to work on.
Whether it's another suburb or another country, you may have staff in different offices who need to work together.
Or perhaps you want your staff to access their work from home, or while they're on the road.
A Virtual Private Network - more commonly abbreviated to "VPN" - uses the internet to connect a remote user to your server.
This connection is encrypted, keeping your data safe from prying eyes.
This is a great way to make existing IT infrastructure available to you and your staff from anywhere in the world.
A VPN connection is great way for you and your staff to access their work from home or on the road.
But - depending on the internet speeds available in your area - if you're using it to connect whole offices together, it might not be realistic to run all your network traffic through your internet connection.
Here, it can make more sense to mirror your servers - to synchronise the files across locations. That way, any projects your staff need to access can be delivered from a local server.
The big trend in small business IT in recent years has been the move to cloud services.
Much of the functions that used to be delivered from a business's server room - such as email, customer relations management, file storage, bookkeeping, word processing, project management and point of sale - are now being run by third-party service providers over the internet.
This means that you and your staff can log in from anywhere you have an internet connection.
If you're building a new business, using cloud services is fairly straightforward.
This can be a bit more complicated if you need to migrate from existing infrastructure. There's the hassle and expense of transferring over your existing systems, and your staff might need to learn new software too.
Do your utmost to avoid a situation where every location has a differently configured network.
These networks should be the same, right down to the colours of the cables used for different parts of the network.
That way, when your IT support goes from location to another, they don't have to spend any time working out how the thing is put together.
This saves you money in support costs. One way or another, it's you who's paying for your technician's time.
There are a few common problems that small businesses keep having to fix. Most of these are not all that difficult to fix.
Every device that connects to your network gets an “IP address”. This is used to work out which bits of data are meant to where. It's not all that different to the address written on the back of an envelope to let the post office know where to deliver the letter.
Most of the time, IP are assigned dynamically - the device joins the network, and gets assigned an available IP address.
Where this can go wrong is that, after a while, the device's IP address expires and it gets assigned a different one.
So when your computer goes to look for the printer, it can't find it.
The solution is to get your technician to assign a static IP address. That means it will always be at the same IP address.
When someone forgets their password, you should ask your network administrator to reset it.
But if this is happening a lot, and you're outsourcing your IT services, being liable for a fee every time might get expensive.
In this instance, you should ask your technician to provision this role to someone trustworthy, such as the operations manager.
Does the network take too long to respond? It's not just frustrating. It's a choke hold on everyone's productivity.
Here are the most common reasons why this happens: