Network attached storage devices – aka “NAS” devices – are commonly sold into small-to-medium enterprises as a solution for storing files on their network. They typically come from a number of manufacturers, commonly QNAP, Synology, Asusstor, and Netgear.
On a business network, there tends to be two types of devices – laptops or desktop computers that people use, and servers. Servers are designed to hold data and do other jobs that don’t require a human being to directly interact with them. Before the cloud become so prevalent in SME IT, it was pretty much required for a network to have a server on it, and the most typical use case was that that server held a copy of the company’s files. It was so common that for a while Microsoft sold a product called Microsoft Small Business Server for just this job.
The discipline that came with this is that we are used to storing files on a “shared drive” or “network share”, and typically laptops and desktops when logged on have a “mapped drive”, and the end result is that we are used to storing files on a computer other than the one that we are sitting at. There are really good reasons to do this, not least is that we a) need to share files with colleagues, and b) need those files backed up. A file server solves both these problems beautifully.
The “pain” that a NAS device solves is that if you believe having a server is a complicated solution to having somewhere you store your company’s files, a NAS device is pitched as a less-complicated solution. Rather than having a complex computer sitting on your network that you have to buy and manage, you buy a NAS device, plug it in, switch it on, and off you go. As in IT “complicated” typically is synonymous with “expensive”, NAS devices are sold on the basis that they are cheaper to buy and cheaper to run than an equivalent server.
This claim may well be generally true, but there are two key problems with NAS devices. Firstly, they are proprietary implementations of an open standard. A Windows file server sitting in an office is, to all intents and purposes, an open implementation. This means you can access it however you want, install whatever backup solution you want, install whatever cloud-synchronisation solution you want, install any virus solution you want, etc. NAS devices do not have this same openness – the NAS vendor looks to deliver value by “locking down” and hiding complexity of the computer behind their own front-end. This is the value proposition of a NAS – it is a server, without the complexity.
This make sense if you are selling a solution based on “it’s less complex” – i.e. you are pitching a simpler solution. But practically all you are doing here is obscuring complexity unnecessarily. If you need a server on the network – and the emphasis here is important – you likely need an open server.
The problem with NAS devices as a proposition to SMEs is the issue of “need”. Most SMEs do require some way to store files so that they can be shared, but in 99 times out of 100, the best place to put files so that they can be shared is in the cloud. Most SMEs do not need a NAS device – what they need is SharePoint or Google Drive, depending on their “Coke v Pepsi” choice of Office 365/Microsoft 365 vs G Suite.
NAS devices come from a time when we aligned the physical boundaries of the office environment with virtual practice – that to all intents of purposes, to do business “things” you had to be in an office, sitting in front of a computer, 9am to 5pm. (This article was written about a month in the UK’s COVID-19 lockdown, and even before then it was clear that we were moving away from constraining business activities within the office and being more open as to where employees did their work – office, home, McDonald’s, who cares so long as the work is done...) Putting data in the cloud makes it available to everyone everywhere (more properly: “every employee every when”), and the reason why we do this is that it is easier to put things in the cloud and let people access it from anywhere than it is to keep things on our network and hook up some way for people to get access to it.
Traditional network design is based on a “castle” paradigm – you have to be inside the walls to access the network, and anyone inside the walls is typically trusted. Because we confer a great deal of trust on people who are inside the network, we make it hard to get “inside the castle”. Whilst every NAS vendor out there tries to fudge this issue, the overall design principle is that to get at data on your NAS, you need to be inside the castle. This is fundamentally backwards when it comes to best practice – you should have a small number of physical devices, and rely on the cloud to store and provide access to your data.
These two issues – that they are network devices designed to live inside a castle’s walls, and that they are closed implementations – makes NAS devices too inflexible for modern business. They are an old approach who’s time has passed.
If you have a NAS device, you should strongly think about migrating off and into cloud storage. If you don’t already have one, in most use cases they are not a good fit. There are some edge cases where they may work for you, but if you are in such an edge case you may well find that a normal Windows or Linux server does what you want without requiring you to pinch off your options so much.