User Manual

> White Paper | Best Practices in Digital Transformation
4
Security
The operators of MTDCs are faced with the tasks of securing
growing and evolving network architectures against increasingly
sophisticated and targeted attacks, while at the same time meeting
ever more stringent compliance and regulatory requirements to
protect the data with which they are entrusted.
This includes threats on the cyber level such as DDoS attacks, web
application attacks such as SQL injection and cross-site scripting,
ransomware (where data is held for ransom supposedly until a
ransom is paid). DNS infrastructure attacks, malware including
Trojans, viruses and worms, phishing, unpatched software, social
media threats and advanced persistent threats (APTs) via (spear)
phishing. Just like natural viruses which mutate in order to survive
against antibiotics, so the list above will mutate and combine in
order to present a more sustained and complex mode of attack.
While the focus of security has moved to cyber-threats based on the
changing profile of the data infrastructure landscape, other sources
of disruption should not be ignored. Hacking, malware and threats
delivered via social media have grown the most to 2015 in terms of
numbers while physical, environmental and disruptions caused by
error and ‘misuse’ have remained at a consistent level. This cluster
of threats remain at a level however where they cannot be ignored.
There are a number of important considerations for developing a
security model in an MTDC:
• It needs to reflect the changing design and operation of a
multi-tenant facility and be designed for the mix of services and
environment oered by the facility (or facilities).
• It must be able to adapt so that it provides consistent, constant
and intelligent protection across evolving and hybrid data center
models.
• It must provide protection against advanced and evolving threats.
• It will only be as strong as its weakest link therefore it needs to
observe principles of ‘absolute’ protection, described as ‘End to
end’, ‘Layered’ or ‘Zero Trust’ depending on the form of security.
It is evident that traditional network-centric security systems based
on perimeters and firewalls are no longer adequate for any but
the most legacy MTDC. Traditional security policies are defined
for security zones that are static and tied to physical devices, and
which are signature-based. Although these have evolved to next-
generation firewalls (NGFWs) that can implement policies based
on applications, users and content, they are still static and rely
on traditional network constructs like IP addresses, virtual local
area networks (VLANs) and server zones. Traditional defenses like
firewalls, IPS, anti-virus and gateways are simply no match for
continually evolving and sophisticated cyber threats, which can
blend malicious techniques.
The MTDC environment is now software-defined, distributed,
interconnected, dynamic and user-focused. The cybersecurity
risk to the service data center is made worse by dependence on
virtualization, cloud computing and the internet of things (IoT).
This creates a complex and dynamic network that gives more
opportunities for attackers to compromise a facility.
It now widely accepted that security must be deployed throughout
an organisation’s data infrastructure, and out to the growing number
of endpoints that are connected to that infrastructure and used to
access cloud services. There also needs to be greater visibility into
the network, as well as enhanced segmentation and control.
There are two key methodologies in securing the next-generation
network and everything connected to it. The first involves reducing
the attack surface and, the second involves mitigating the risk.
However, there are numerous dierent - but often parallel -
approaches to both.
Reducing the attack surface aims to prevent all but authorised
access to system assets, and to establish access rights. This can
be achieved by minimising exposed system targets, controlling
system and network segment access across the network, enforcing
least privilege for all security subjects, or reducing the amount of
data that needs to be scanned by deploying trusted software and
procedures.
In order to mitigate the risk of attacks, the MTDC provider must
also understand the risks and then implement specific measures to
reduce or minimise unacceptable risk, such as those to reduce the
severity of the consequences of an attack, reduce the probability
of an attack occurring, or reducing exposure to the attack. This
might include trac segmentation to filter and verify network
activity to reduce the potential attack surface include firewalls and
switch access control lists (ACLs), as well as the creation of subnets
and logical segmentation for internal trac and gain visibility into
applications, users and content.
Security must work across the multi-tenant data center so that
client organisations can rest assured that their data and applications
are safe. This means a consistent and accurate security policy across
heterogeneous environments. The programmability aspect of new
security technologies is all software-based. Indeed the majority rely
solely or to a large extent on software.
Software-Defined Security is adaptive in that the security policy and
controls automatically remain with the device if is moved, migrated
or scaled, which speeds up response time and reduces the scope
for human error.
Increasingly, sophisticated algorithm-based techniques are used,
not just to identify security threats but to diagnose the wider
principle of ‘data health’. There are two main aspects to threat
intelligence: technology or machine intelligence, and human
intelligence. While machine intelligence is able to mine and analyse u