12.3 Introduction to File Sharing
Your Linux machine probably doesn’t live alone on your network, and when you have multiple machines on a network, there’s nearly always a reason to share files among them. For the remainder of this chapter, we’ll first look at file sharing between Windows and macOS machines, and you’ll learn more about how Linux adapts to interacting with completely foreign environments. For the purpose of sharing files between Linux machines or accessing files from a Network Area Storage (NAS) device, we’ll wrap up by talking about using SSHFS and the Network File System (NFS) as a client.
12.3.1 File Sharing Usage and Performance
One thing you need to ask yourself when working with any kind of file sharing system is why you’re doing it in the first place. In traditional Unix-based networks, there were two major reasons: convenience and lack of local storage. One user could log in to one of several machines on a network, each with access to the user’s home directory. It was far more economical to concentrate storage on a small number of centralized servers than to buy and maintain a lot of local storage for every machine on the network.
This model’s advantages are overshadowed by one major disadvantage that has remained constant over the years: network storage performance is often poor compared to local storage. Some kinds of data access are okay; for example, contemporary hardware and networks have no problems streaming video and audio data from a server to a media player in part because the data access pattern is very predictable. A server sending the data from a large file or stream can pre-load and buffer the data efficiently, because it knows that the client will likely access data sequentially.
However, if you’re doing more complex manipulation or accessing many different files at once, you’ll find your CPU waiting on the network more often than not. Latency is one of the primary culprits. This is the time it takes to receive data from any random (arbitrary) network file access. Before sending any data to the client, the server must accept and decipher the request, and then locate and load the data. The first steps are often the slowest, and are done for almost every new file access.
The moral of the story is that when you start thinking about network file sharing, ask yourself why you’re doing it. If it’s for large amounts of data not requiring frequent random access, you likely won’t have a problem. But if, for example, you’re editing video or developing a software system of any substantial size, you’ll want to keep all of your files on local storage.
12.3.2 File Sharing Security
Traditionally, security in file sharing protocols has not been treated as a high priority. This has consequences for how and where you want to implement file sharing. If you have any reason to doubt the security of the network(s) p. 324 between the machines sharing files, you’ll want to consider both authorization/authentication and encryption in your configuration. Good authorization and authentication means that only parties with the correct credentials have access to files (and that the server is who it claims to be), and encryption ensures that no one will be able steal file data as it transits to its destination.
The file sharing options that are the easiest to configure are typically the least secure, and unfortunately, there are no standardized ways to secure these types of access. However, if you’re willing to put in the work of connecting the correct pieces, tools such as stunnel, IPSec, and VPNs can secure the layers below basic file sharing protocols.