Or why users put data where they do.
I’m going to start this time with a caveat: what follows is a discussion piece about an emotive and complex area, with a lot of similar but not identical challenges. As with a lot of technology, there are different options available to solve real problems; to get the best results it’s really important to understand what you want from your technology – your requirements – and select the solution(s) best suited to them, and that includes non-technical considerations such as cost. Best suited does not always mean perfectly suited.
One of the more significant obstacles to cloud adoption is being able to get users to move their data to cloud storage. There can be a few reasons behind this reluctance, including:
- Users are unsure whether their data is being backed up
- Nervousness around putting data in a public cloud
- The data becomes disconnected from the application that produced it
- Shared data spaces are very different from what they’re used to (e.g. SharePoint vs. file share)
In education, users are lucky to have access to services such as Office 365 and G Suite for free, and of course that gives them more free storage in either OneDrive or Google Drive than they can shake a stick at.
I’ll put my data in OneDrive or Google Drive then…
Well, yes, you could do, and in a number of circumstances that might be the right thing to do. And if you did it would make for a very short blog.
The thing to remember about Office 365 and G Suite is that they are really designed for collaboration and personal document management; they’re not really designed to directly replace the general-purpose file share. They work really well for workgroup or team activities such as sharing documents or simultaneously updating a document with a few people, but not so much when it comes to storing very large files or making data accessible to applications that are not ‘cloud aware’.
The rule of thumb with data is that it’s best to keep it close to the application. Both Office 365 and G Suite are designed for easy access to data held in cloud storage. If you’re using Word Online or Google Docs, accessing your files stored in OneDrive and Google Drive respectively is pretty straight forward. Even using a desktop version of Microsoft Office and accessing data in OneDrive is pretty straight forward – Office is ‘cloud aware’.
The Human Factor
For those of us who have been involved with technology for some time, we’ve become used to a certain way of doing things. We login to the local network, go about our daily tasks and sometimes even produce some outputs in the form of data. If we don’t need to make it accessible to others we save that data in our network home folder – because saving it locally on a device is a bad thing, right? – and if we do need to give others access to the source file, we save it in a shared area. We save to our home folder or shared areas because they’re held on a server, and frankly our network admin is much more reliable when it comes to backing our data up than we are.
As the Software as a Service (SaaS) solution(s) available to us becomes more embedded, we start to store some of that data in cloud storage. But we don’t necessarily store all of it there. Why not? Well, we humans are an adaptable bunch and we tend to follow a ‘path of least resistance’ approach – we do what’s easiest for us. For example, if you’re producing a lengthy video using a desktop application – inherently a large data file – it’s unlikely to be a pleasant experience to save each edit directly to cloud storage. You might save the final, finished version there, but while you’re still working on it uploading a large file on a contended and comparatively low bandwidth connection is going to take a while. As performance has increased we’ve become less patient – we expect things to happen almost instantaneously these days – and we get a bit irritated when confronted with the rotating circle as we wait for the save to happen. It’s almost as frustrating as waiting for your favourite utility supplier to answer the phone. Or perhaps that’s just me?
So instead we save it locally. Sometimes that means locally on the device, but given that we’ve all learned the lesson the hard way at some point, we tend to take the balanced approach and save it to a network share, which gives us the right mix of performance and protection from data loss.
We may not realise it, but our under-appreciated network admins have put some things in place to make our lives easier when it comes to saving data to network locations.
There are still A LOT of desktop applications in use in UK schools; when we deliver services to create new workstation images it’s common for us to handle deployment of at least 50 applications. At the more extreme end of the scale, it’s pushing 150 to 200 applications. The proportion of those applications that are ‘cloud aware’ – can save directly to OneDrive without the help of a sync client or some other mechanism – is still very low. Actually, most applications are not even network aware.
Since the early days of networks, a common solution to the problem of an application not being network aware has been to simulate a local partition – map a drive to a network share that looks to the application like a drive on the local computer. With CC4, users see this with their home folder being presented as the N: drive, and a general shared area presented as the W: drive.
Connecting to Cloud Storage
So when we, as users, start to adopt cloud storage more there are a few options available to us, and to network managers, to help.
- Use the (OneDrive App) integration built into the operating system to access OneDrive as a library. That gives you access to your cloud files locally, but crucially, not as a mapped drive. And like the sync client, your mileage may vary if the device is used by multiple users.
- Access the data directly in the cloud (ordinarily via a browser). This can result in having to download individual files to work on, then upload the modified version back to cloud storage.
- Install a sync client, such as the OneDrive App, or Google Backup & Sync. Like the OS integrated functionality these keep a copy of the data in a local folder, and then synchronise that data to cloud storage. OneDrive Files On-Demand, which is a new feature in Windows 10 v1709, enhances that functionality by presenting data only held in OneDrive as available in the local folder, and downloads it when accessed. Sync clients are great if you’re the only one who ever uses the device, but are often not viable if the device is shared by multiple users (as student devices in schools tend to be).
- Map a drive directly to the cloud storage. This is possible (cf. CC4 OneDrive Mapper), but can be quite challenging to achieve, as well as challenging to maintain persistently. Because of this, solutions tend to be scripted or programmatical and carry an additional cost.
The Twilight Hybrid Zone
Once the transition has been made to a cloud-first environment it’s likely these problems will go away. The reason they’re a real issue today is that we are in the process of moving from the traditional way of doing things, to the modern. We have some components, such as desktop applications, that work in the traditional way and others that are designed to fit with the modern approach – we’re in the hybrid zone.
The demand for hybrid is not likely to go away anytime soon. There was a subtle but detectable shift at Microsoft Ignite this year, with a number of developments aimed at solving problems in the hybrid space. So it looks like Microsoft don’t think it’s going away anytime soon either.
One of those developments pertinent to data storage location is Azure File Sync, which we’ll look at in the next instalment.