Integration

Hi Friends! In this post we are going to make a brief summary of the most important thinks that we explained in our blog. Remember that for further detail you would need to read these posts, but in order to explain some integration of the parts that compose our designed data center, we are going to talk about: network, security, servers, management, QoS, facilities and storage. Let’s go!

Network: Main solutions on today’s data centers include SDN, as it is a very useful solution to implement virtualization and cloud. SDN is a software defined network can be managed from a single console, so the devices do not have to be configured individually. The network administrator can make the necessary changes simply by modifying the network software. Independently from this SDN importance on our project, we want to say that we will use the Spine-leaf topology, as it’s one of the most common used nowadays and connects all the devices between them.

Also, it’s very important in the network section the role of the virtualization; as if we invest in virtualization the number of physical devices that our network needs will descend considerably, in second place for the growth of failures tolerance of our network, once we establish a virtualized model from our network, many failures can be handled through software and avoided, in third place this practice will add an extra layer of security to our network.

Security: We started our design thinking on the specifications that the security from our data center has to accomplish, and then we create a security policy. Based on physical and logical security:

Physical security:

1. Place your network layers into different spaces
2. Be aware of surroundings
3. Train your employees
4. Perform regular internal audits
5. Lock everything

Logical security:

1. Unified management console
2. Scan for applications and vulnerabilities
3. Coordinate communication between security devices for visibility into data flows
4. Choose security technologies that are virtualization-aware or enabled

Servers: All the part from our Data Center servers is based on virtualization, specifically with the use of software IBM Power Systems servers and PowerVM servers. This system allows us to do several things:

1. Take up to a thousand virtual machines on a single physical machine
2. Easy management
3. Create, suspend, activate or delete virtual machines in a simple way
4. Resource sharing
5. Movement of services between several machines
6. Provisioning resources dynamically

Management: We elaborated a very retailed list about which point to consider when preparing the management of a data center. Those are a collection of points which will help you to organize your work:

1. Software/Hardware management
2. Dependencies tree control: Decide and set which systems depend on which others.
3. Device monitoring: Keep information about all the devices in a log database.
4. Security management: Running backups and testing the security of your network
5. Device discovery
6. User management: establish levels of access to your employees.
7. Network performance analysis
8. Network perimeter intelligent adjustment: segregate your network on several parts and analyse the traffic flow in and out to detect anomalies
9. Standards management: segregate your network on several parts and analyse the traffic flow in and out to detect anomalies
10. Intelligent notifications

QoS: We will center our attention in the bandwidth and the flexibility to outfit our applications with the best properties we can. The requirements for our QoS include:

1. Adapter-based QoS with hardware protocol offloads to avoid using server CPU and RAM resources
2. The ability to assign applications to virtual connections with a guaranteed amount of bandwidth
3. The ability to allow applications to burst above their guaranteed amount if extra bandwidth is available