IaaS Security risk Resolution 6 recommendations

IaaS provides users with computing, storage, networking, and other basic computing resources on which users can deploy and run arbitrary software, including operating systems and applications, without the user managing and controlling the underlying infrastructure, but controlling the operating system, storing, deploying applications, and having network components such as host firewalls The ability to have limited control over permissions.

1, customer data controllable and data isolation.

For data leakage risk, the resolution of such risks is primarily through data isolation. Data isolation can be achieved through three types of approaches:

First, let customers control the network policies and security they need to use.

Second, from the storage side, the customer's data should be stored in the virtual device, because the virtual storage is in the larger storage array, so the virtual store, you can at the bottom of the data isolation, to ensure that each customer can only see their own corresponding data.

Third, in the virtualization technology implementation, you can consider large-scale deployment of virtual machines for better isolation, as well as the use of virtual storage file systems, such as VMware's Vmfs file system.

2, comprehensive consideration of data center hardware and software deployment.

In the hardware selection, consider brand manufacturers, the choice of hardware to take into account the quality, brand, ease of use, price, high maintainability and other factors, and select cost-effective manufacturer products. In the choice of virtualization software, we also need to balance the price, the manufacturer and the quality. It is recommended that customers who have the conditions first employ consulting firms for consultation. On the market, there are now three cloud computing alliances that can provide a complete cloud computing solution, ranging from the underlying hardware to the top software. respectively, Cisco+emc+vmware,ibm,hp+microsoft.

3, establish a secure remote management mechanism.

By definition, IaaS resources are at the far end, so you need some remote management mechanisms, and the most common remote management mechanisms include:

VPN: Provides a secure connection to the IaaS resource.

Remote Desktop, remote shell: The most common solution is ssh.

Web Console UI: Provides a custom remote management interface, typically a custom interface developed by a cloud service provider, such as the Rightscale interface for managing Amazon AWS Services.

The corresponding security policy is as follows:

A. The best way to mitigate the threat of authentication is to use two-factor authentication, or to use a dynamic shared key, or to shorten the shared duration of a shared key.

B. Do not rely on reusable user names and passwords.

C. Make sure the security patches are in time.

D. For the following programs: SSH: Authentication using RSA keys; Microsoft Remote Desktop: Use strong encryption, and require server authentication; VNC: Running it on an SSH or TLS tunnel; Telnet: Do not use it, if you must use it, it is best to use VPN.

E. For programs that cannot protect the transmission of data, you should use a VPN or secure tunnel (TLS or SSH), first use IPSec, then SSLv3 or TLSV1.

4, the choice of safe virtualization vendors and mature technology.

The best choice is to have sustained support and long-term security concerns of vendors. Periodically update virtualization security patches and focus on virtualization security.

Mature virtualization Technology not only can prevent risk, in large cases can also enhance system security, such as VMware on the problem of virtual machine isolation, DRS system dynamic scheduling, Microsoft Hyper-V technology to prevent unauthorized virtual machine communication between.

5, establish and improve IT industry regulations.

In a cloud computing environment, users don't know where their data is, so there's a certain amount of anxiety, such as where my data is, safe, and so on.

In the IAAS environment, because the virtual machine has the drift characteristic, the user largely does not know that the data actually resides on that server, the storage. In addition, because of the unique characteristics of the data, once known to others, the value will be drastically reduced.

This needs to be regulated from two angles of law and technology, firstly, to establish and perfect the law, to strictly restrain the immoral behavior of the IT employees, and to prevent the unsafe phenomena such as data leakage from the angle of human. Second, the development of virtual machine drift tracking technology, IaaS under the unique data encryption technology, so that users can track their own data, sensing the security of data storage.

6, for the sudden interruption of service and other irresistible new factors, to take the three central strategy of the two places.

Risks such as service outages exist in any IT environment, and when deploying cloud computing data centers, it is best to take a strategy based on the three centers of the two places to back up data and the environment. As shown in the following figure.

Production Center and city Disaster Preparedness Center in the same city, distance within 10Km, off-site disaster preparedness Center is usually in another part of the country, distance can span several provinces.

Production Center for the external Service of the main center, as with the city Disaster preparedness Center near the distance, you can take the bare fiber-optic connection, take the synchronous replication mode, real-time data to maintain synchronization. City Disaster Preparedness Center and Remote Disaster Preparedness Center data due to distance, only WAN connections can be taken, so asynchronous replication mode is taken.

In this environment, once the production center has been destroyed, the City disaster Preparedness Center can undertake the task of external services in real time, in this case, users do not feel any interruption of service. In the event of a large area of destruction, such as an earthquake or war, production center and the same city disaster Preparedness Center service at the same time interruption, can enable off-site disaster preparedness Center external services, in this case, because the data needs to be restored, users feel the service interruption, but in a short period of time will be restored, will not cause serious accidents.