Adopting IaaS: Tips and Best Practices for Cloud Transformation
Many organizations are choosing to adopt cloud and hybrid cloud architectures to integrate with infrastructure-as-a-service (IaaS) solutions. It’s easy to see why, given the many benefits and advantages. These include:
- Flexibility to pay for only what is used and provisioned;
- Economy of scale, which enables sharing of the investments across branches;
- Vendor-provided, cost-effective and efficient IT maintenance and operation; and
- Increased speed for faster innovation.
An effective cloud transformation requires engagement from all stakeholders across the organization. It is important to consider the company’s overall culture and security posture during the implementation process.
Regarding security, expectations vary depending on the type of cloud the organization adopts. If it chooses to focus on IaaS, its main objective is to cut IT expenses and complexity without sacrificing security in the infrastructure or access processes. With a platform-as-a-service (PaaS) solution, however, the focus shifts to securing applications and data, as well as being in compliance with regulations.
Security should always be on the table whenever an organization decides to transform its infrastructure or adopt a new service, from the design phase to the final implementation. Cloud solutions introduce many advantages, but with these advantages come more complex security concerns.
When building a service, it’s important to consider all security controls to reduce the risk and increase the efficiency of each element. This means securing infrastructure and applications, managing identities and access, and securing all elements involved in the execution of a transaction. This also applies to services that have elements stored in the cloud. If a cloud service provider supplies the infrastructure, for example, it’s important to ensure the vendor will also provide the security controls.
Of course, this type of infrastructure can impact the service-level agreement and cause customers to demand transparency. If the infrastructure supplied by the cloud service provider comes under attack, for example, it’s important to alert the customer’s security information and event management (SIEM) provider.
About the Security Controls
Just because the application or database installed on the IaaS is secure doesn’t mean the overall service is secure. It’s usually ideal to have a single database protection capability with visibility and control over the database, cloud and on-premises tools.
The security controls in an effective IaaS program should include the ability to:
- Manage data center identities and access.
- Authenticate, authorize and manage users.
- Secure and isolate virtual machines (VM).
- Patch default images for compliance.
- Monitor logs on all resources.
- Isolate networks.
Failing to check any one of the above boxes means compromising the security of workloads moved onto the IaaS system. An effective IaaS provides customers with a multilayer security strategy.
Managing Data Center Identities and Access
IBM SoftLayer, for example, employs security staff to monitor its access sites 24/7. Access to the data center requires an ID badge and biometric authentication. Then each level of the data center requires its own set of credentials, and only staff members whose roles require access will be permitted to enter a given level. All access is logged and monitored by closed-circuit television.
Authentication and Authorization
An effective IaaS solution includes policies that enable clients to create and manage user accounts and assign privileges. It should be able to check source IP addresses, prevent users from accessing the portal, and monitor activity to implement and effectively manage an access policy.
IaaS provides user management and granular access/permissions capabilities for elements provided by the platform, including servers, storage and networks. Many solutions rely on the client to create and delete portal users. If the client is managing the servers, the service provider would defer to the client for this process. If the provider is managing the servers, it should work with the client’s team to ensure the proper user accounts and privileges are available to the correct personnel.
What are the risks associated with insecure VM isolation? What are the consequences of isolation failure?
In short, they are multiple and serious: Cybercriminals can leverage weak VM isolation to manipulate assets inside the cloud IaaS. In a VM hopping attack, for example, bad actors compromise one VM to gain access to the other VMs located on the same hypervisor. The attackers use this access to switch off the system, compromise data and replicate multiple VMs to jack up the cost to the customer.
Patch management involves patching shared devices, such as switches and routers, within a period consistent with security best practices. A highly automated cloud environment can accomplish patching by the time a new compliant server or workload migration starts up.
An IaaS solution should provide monitoring and incident management services. It should also establish explicit policies and processes for logging security events. Logging capabilities must include:
- Ongoing monitoring and management;
- Monitoring of network traffic using various techniques; and
- Analysis of security logs generated from the platform component related to irregular, suspicious activities.
Logged alerts should be handled in a timely manner and, if applicable, communicated to clients. The support and incident response teams should notify clients of any activity related to the infrastructure.
IaaS solutions often use firewalls to control internet access to VMs. These capabilities enable clients to build secure, internet-facing environments, support shared and dedicated firewalls, and supply additional network controls.