Why it’s time for manufacturers to take security in the cloud seriously
Manufacturers deal with sensitive data every day. This includes test and quality data, warranty information, device history records, and especially the engineering specifications for a product that are highly confidential. Trusting that data to a cloud-based application or cloud services provider is a major step, and manufacturers need to fully educate themselves about the security risks and advantages of cloud-based software.
As we prepare to enter the second half of 2018, consider the following three questions as your guide when discussing application infrastructure and operations with cloud providers.
Question #1: How do you keep my data safe?
The answer should be long and multi-faceted. Because no single tool will defend against every kind of attack in any network, cloud providers must deploy multiple layers of defense using: internal systems; protection provided by tier 1 cloud platforms; and security service providers. All of these elements come together to provide complete protection.
Below are some examples of these layers:
- Physical defence: Cloud platform providers can and should exercise tight control of access to the physical devices on which the software systems reside. In best case scenarios, Independent auditors attest to the safety of this access. This control and documentation must be reviewed on a regular basis.
- Barriers to entry: Firewalls built into the cloud service can limit access to ports managed by the application. Unneeded ports should be blocked so that they cannot be accessed.
- Application password protection: the best-designed cloud applications allow your organisation’s identity management system to provide authentication and password management, limiting access to your data and following your internal security policies. This should also support two-factor authentication if your internal policies require it. Some of the more advanced systems can also provide an identity management service as an alternative to your internal solutions, if required.
- Application firewalls: Most enterprise-class application designs will include a Web Application Firewall service that uses the latest technology to defend against such things as denial of service attacks and other types of malicious access.
- Activity monitoring: State-of-the-art cloud platform providers continuously monitor for suspicious activity that could be the result of hacking or malware. Again, in best case scenarios, warnings are sent automatically and steps taken to protect the data and the integrity of the platform.
- Malware monitoring: Both the application provider and the hosting platform provider must run active checks for malicious code to ensure each piece of code that is executed matches the published signature for that code. Be warned: this is a step that many providers have not migrated to yet.
- Code standards: Good security starts with good code. Security standards must be included in the system development life cycle, governing every aspect of the system. Be sure to review the code standards of the application developer.
- Third party code scanning: The most advanced application providers use a third-party firm to scan code looking for opportunities to improve security and look for known vulnerabilities with each new version of the application. Ask for details about this, as there are many different levels of scanning available; a once-a-year scan is obviously not as valuable as regularly scheduled scans before each new release of software.
- Data encryption: Generally accepted practices for data encryption provide different options for data in different modes: data in transit (being communicated within the system or between the database and your user interface) and data at rest (data that resides within the database and is not currently being accessed).
Data in transit can be encrypted using industry standard encryption through the browser. Additionally, APIs that access the data should use encrypted data and include encrypted tokens to increase access control.
Encryption of data at rest protects against accessing data from outside the application’s control. As the physical access to the system is protected and the data is in password protected databases, at-rest encryption may not be essential for every customer - but the question is still worth asking.
Question #2: How do I know that my data can’t be accessed by other customers?
There are many ways to ask this question:
- Do you mix my data with other companies’ data?
- Can other people see my data?
- What’s your data structure for each customer?
The answer to each of these is data separation. The system architecture should ensure separation of customer data by customer organization, usually by individual factory or site. This allows even customer administrative tasks such as assigning roles to be limited in scope. While many applications are multi-tenant (meaning the application is shared across multiple customers), transactional data should still be separated by customer factory, meaning there is no commingling of customer data. In other words, your data will be separated from every other customer, giving the highest level of data separation.
Question #3: How does cloud security compare to on-premise security?
There is a common misperception that a set of servers running on-premise at a corporate office is more secure than a cloud-based application. Owning the hardware and software often gives a false sense of security; most on-premise systems fall far short of the security that the best cloud providers have deployed.
For example, the cloud storage system utilized by my company was designed for 99.999999999% durability and up to 99.99% availability of objects over a given year. That design and those numbers are virtually impossible to duplicate with an on premise solution. In addition, the comprehensive access control described above is nearly impossible to duplicate on-premise. To deploy tools like these in an on-premise environment would require not only large investments in infrastructure, but large teams to manage them too.
- How big is your security team?
- How much is your budget for security around your manufacturing data?
Then remember, the best application providers and data centers have large, dedicated security teams who have implemented automated threat monitoring systems that operate 24x7. In the end, the best cloud software companies have dedicated more time, resources and budget to securing our systems than most organizations are able to provide themselves.
- » Google Cloud secures support for NVIDIA’s Tesla P4 GPUs with more machine learning goodness
- » How cryptomining is the attack vector du jour - as hackers increasingly target cloud infrastructure
- » Google and Alibaba focus on Southeast Asia in latest infrastructure expansion
- » The cloud skills gap can be tamed – as an example from telecoms shows
- » The data centre of tomorrow: How the cloud impacts on data centre architectures