The Security in Big Data?

258 Views
Oct 11, 2021
The Security in Big Data?

In the era of all-out digitalization, attacks and data theft regularly make the headlines. And if the stakes are just as real as the awareness is massive, including from the general public, for companies, fingers crossed is not enough. If data is everywhere, so must security.

Deperimetrization: the neologism is launched, as barbaric as it is abysmal. It indicates a decompartmentalization of the information system and without a precise scope, the data can be created, captured, processed, modified, stored anywhere, anytime.

De-erimeterization therefore offers a gigantic world of possibilities, aided by rich (structured, unstructured, multimedia, etc.), polymorphic and multi-source data, with exponential growth further accelerated by connected objects, from the smallest network sensor to the smart and connected car.

Be careful, however, behind this apparent “nebula” of data, physical media remain (including wireless ones, because although using the principle of radio waves, immaterial, they remain physical media). In this context, like matter, it is possible to characterize data according to three states: data at rest (solid state) on storage disks, data in transit (liquid state) on networks, and data being processed (gaseous state) on servers.

This educational analogy makes it possible to understand the link between data, their physical media according to their respective states and the security that must result from it, over the entire data lifecycle.

Secure the data, including during its processing

Recognized as a real asset of the company for a long time, data benefits from special attention in terms of its protection. In terms of storage, its confidentiality and integrity are most often ensured by access controls (physical and logical), encryption and other partitioning mechanisms between disks, virtual machines, processes, etc.

Likewise, protocol security (IPSEC over IP, TLS over http or WPA2 or 3 over WiFi, etc.) ensures data protection when it is in transit. With particular attention paid to wireless networks, it is so easy to “listen” to streams that are not protected there.

When it is being processed, the data is also particularly vulnerable. Its protection is all the more difficult to understand in this context as it also involves the programs responsible for its treatment. The risk being that a malicious modification of the program could lead to an action which would not have to be; for example, a vehicle that does not brake when it should (or vice versa).

Therefore, it is the entire integrity of the program, its configuration and its intelligence (IA) that must be preserved to ensure the security of the data.

IoT security, the first bulwark

With the proliferation of digital devices and the connectivity brought to objects that were not previously, such as vehicles, refrigerators, and even production lines, data is therefore everywhere now: it is even around these objects that data is often captured and digitized, and sometimes restored.

External to the information system by definition, connected objects also represent a colossal attack surface, with as many potential entry points to organizational data. They are therefore the first elements to be secured in order to limit the risks of malicious acts, in particular in strategic sectors, impacting human health or with particularly important economic issues.

This is obviously the case for the automobile, medical (connected health), large networks and Operators of Vital Importance (water, electricity, telecommunications, army / defense), and all major industries including production stoppages. are economically dramatic.

For all these activities in particular, but also for the entire economy, only an integrated and homogeneous level of security from end to end, from connected objects to the depths of the Cloud, will ensure the security of a world where data is everywhere.

The challenge today is therefore to think of unified security with coordinated policies (keying, maintenance, reaction in the event of an attack) and technologies correlated with one another, in order to obtain homogeneous levels of assurance, whatever the state of the data (solid / liquid / gas).

A security that must also benefit from better orchestration, in the face of a security value chain today very fragmented and the absence to date of a player in charge of the specific profession of security operator, who would supervise this value chain and assume its liability.

Finally, and in a context of sovereignty, technologies allowing systemic security management must emerge. The context is favorable (EU cyber act) and the strategic vision now understood of the ecosystem. It’s up to us to act!

Leave a Reply

Your email address will not be published. Required fields are marked *