Storing your data – Cloud vs. Local Storage for Imaging Systems

Anyone working in imaging knows that one major ongoing problem is handling the massive amounts of data collected. A relatively low resolution camera can output massive volumes of data, if combined with a stage, z motor, wavelength switching hardware or just run in video collection mode. Sooner or later, many users end up needing a storage solution beyond a laptop or carry-around hard drive.

Two typical solutions exist: Cloud Storage, and Local Storage. Now, everyone hears “cloud”, and thinks google drive or similar backup services, but the term cloud has also come to mean any storage which isn’t inside your office. To this end, one could consider the IT Department’s servers at a university to be “cloud” storage.

Cloud Storage

The advantage of storing data offsite are compelling. They usually revolve around the following points:

• Offsite physical security (If the lab floods your data is safe)
• Low upfront cost, as you aren’t purchasing a physical system.
• Lack of required maintenance
• Potential for massive volumes of storage
• People outside of the facility can access cloud data easily

Depending on your facility, the costs for on site storage can range wildly. If you want a target comparison price to consider,  I’d look at Goggle, which as of this writing is a cost of roughly $0.10 per Gigabyte per Month. Current Google prices can be found here. This price is not inclusive of the internet connection cost needed to upload and download whatever data you send to the cloud, so that must be considered as well.

The downside(s) to cloud storage are intrinsic to the advantages provided by offsite systems: If your data is on the cloud, it has to get there, and back to you when needed. This means the time and cost required to upload/download. For many users of complex systems, the size of one data set would take too long to move around between a lab and the cloud. So, it’s a good solution, but not perfect. In the case of inside-facility cloud storage, some of these transfer problems can be avoided, as the network speed inside of a properly equipped facility should be hundreds of times faster than the WAN connection speed, not to mention a quality IT department can establish multi-link connections between a server and a workstation to reduce transfer times.

The other consideration for cloud storage is security. Your data on the cloud is accessed by username and password. Can someone get that information? Can they obtain precious data without you knowing? For some, this risk alone precludes cloud storage for their work.

Local Storage

Local Storage is a good option for several reasons:

• Your facility doesn’t have room on the IT server site for your data
• Size of data precludes offsite upload/download, and you need fast local access
• To ensure local physical security for sensitive data
• Ongoing cost of cloud storage and network connection is too high

In the local storage realm, the most common system found today is the “NAS” server. A NAS, or “Network Attached Storage” server, is basically a computer which has one function, to store, and to serve, data to the network. These systems can be set up to work sitting next to an imaging system, basically using them as a massive hard drive, or can be configured for access on a lab-wide network. At home I use a Synology DiskStation. This system provides redundant backup, serves both work and personal files for every computer and device on my network, and does so at a relatively low cost ($500-$700 complete). My system is considered a “Personal NAS”, meaning it’s not really meant for a corporate environment, but systems like this are. These systems are a relatively low-cost option for storing large volumes of data, are physically safe if set up correctly, and quite reliable. Consider that for 1TB of Google space, you’d pay roughly $100 a month. For a $700 investment, my NAS has 2TB of storage, so in roughly 5-6 months, I’ll have recaptured the investment on my NAS. (Not considering ongoing replacement cost of the Hard Drives).  The downside is that you have to administer them, meaning setting one up, buying good hard drives for the unit, setting up how much security vs. exposure is used and managing who gets room on the server. Not to mention physical security – if my house or your lab burns down, my data goes with it.

So which option is right for a typical imaging lab? It really depends on your IT department’s policies, the bandwidth of your internet connection, the type of imaging work you are doing and the cost considerations for your organization. At minimum, it doesn’t hurt to consider the options available before choosing where your data should live.

-Austin