it is the presentation which shows the various topics of the loud computing

Editor's Notes

• #2: Change image
• #14: Virtual layer is deployed on the physical layer. It specifies the entities that operate at this layer, such as virtualization software, resource pools, and virtual resources. The key function of this layer is to abstract physical resources, such as compute, storage, and network, and making them appear as virtual resources. Other key functions of this layer include executing the requests generated by control and forwarding requests to the physical layer to get them executed. Examples of requests generated by control layers include creating pools of resources and creating virtual resources.
• #15: Virtualization refers to the logical abstraction of physical resources, such as compute, network, and storage that enables a single hardware resource to support multiple concurrent instances of systems or multiple hardware resources to support single instance of system. For example, a single disk drive can be partitioned and presented as multiple disk drives to a compute system. Similarly, multiple disk drives can be concatenated and presented as a single disk drive to a compute system. With virtualization, it is also possible to make a resource appear larger or smaller than it actually is. Further, the abstraction of physical resources due to virtualization enables multitenant environment, which improves utilization of the physical resources.
• #16: Virtualization when deployed offers several benefits to build a cloud infrastructure. It enables consolidation of IT resources that helps service providers to optimize their utilization of infrastructure resource. Improving the utilization of IT assets can help service providers to reduce the costs associated with the purchasing of a new hardware. It also reduces space and energy costs associated with maintaining the resources. Moreover, less people are required to administer these resources, which further lower the cost. Virtual resources are created using software that enables service providers to deploy infrastructure faster as compared to deploying physical resources. Virtualization increases flexibility by allowing to create and reclaim the logical resources that are based on business requirements.
• #17: While building cloud infrastructure, virtual layer is deployed on physical layer. This layer enables fulfilling two key characteristics of cloud infrastructure: resource pooling and rapid elasticity. Virtual layer specifies the entities that operate at this layer, such as virtualization software, resource pools, and virtual resources. Virtual layer is built by deploying virtualization software on compute systems, network devices, and storage devices.
• #18: The software used for compute virtualization is known as the hypervisor. The hypervisor is a software that is installed on a compute system and enables multiple operating systems to run concurrently on a physical compute system. The hypervisor along with hypervisor management software (also known as control software, which is discussed in ‘Control Layer’ unit) is the fundamental component for deploying software defined compute environment. The hypervisor abstracts the physical compute hardware to create multiple virtual machine, which to the operating systems look and behave like physical compute systems. The hypervisor provides standardized hardware resources, such as processor, memory, network, and disk to all the virtual machines. A hypervisor has two key components: kernel and virtual machine manager (VMM). A hypervisor kernel provides the same functionality like the kernel of any other operating system, including process creation, file system management, and process scheduling. It is designed and optimized to run multiple virtual machines concurrently. A VMM abstracts hardware and appears as a physical compute system with processor, memory, I/O devices, and other components that are essential for operating systems and applications to run. Each virtual machine is assigned a VMM that gets a share of the processor, memory, I/O devices, and storage from the physical compute system to successfully run the virtual machine.
• #19: Hypervisors can be categorized into two types: bare-metal hypervisor and hosted hypervisor. A bare-metal hypervisor is directly installed on the hardware. It has direct access to the hardware resources of the compute system. Therefore, it is more efficient than a hosted hypervisor. However, this type of hypervisor may have limited device drivers built-in. Therefore, hardware certified by the hypervisor vendor is usually required to run bare-metal hypervisors. A bare-metal hypervisor is designed for enterprise data centers and cloud infrastructure. It also supports advanced capabilities such as resource management, high availability, security, and so on. In contrast to a bare-metal hypervisor, a hosted hypervisor is installed as an application on an operating system. In this approach, the hypervisor does not have direct access to the hardware and all requests must pass through the operating system running on the physical compute system. Hosted hypervisors are compatible with all the devices that are supported by the operating system on which it is installed. Using this type of hypervisor adds overhead compared to a bare-metal hypervisor, because there are many services and processes running on an operating system that are consuming compute system resources. Therefore, a hosted hypervisor is most suitable for development, testing, and training purposes.
• #20: The network virtualization software is either built into the operating environment of a network device, installed on an independent compute system (discussed in ‘Control Layer’ unit) or available as hypervisor’s capability. The network virtualization software abstracts physical network resources to create virtual resources such as virtual LANs or virtual SANs. The network virtualization software built into the network device operating environment has the ability to abstract the physical network. It has the ability to divide a physical network into multiple virtual networks, such as virtual LANs and virtual SANs. The network virtualization software installed on an independent compute system is the fundamental component for deploying software defined network environment. This software provides a single control point to the entire network infrastructure enabling automated and policy based network management. The network virtualization can also be available as hypervisor’s capability, which emulates network connectivity among VMs on a physical compute system. This software enables creating virtual switches that appears to the VM as physical switches.
• #21: The storage virtualization software is either built into the operating environment of a storage device, installed on an independent compute system (discussed in ‘Control Layer’ unit), or available as hypervisor’s capability. The storage virtualization software abstracts physical storage resources to create virtual resources, such as virtual volumes or virtual arrays. The storage virtualization software built into the array operating environment has the ability to pool and abstract the physical storage devices and present it as a logical storage. The storage virtualization software installed on an independent compute system is the fundamental component for deploying software defined storage environment. The software has the ability to pool and abstract the existing physical storage devices and present it as an open storage platform. With the help of control software (discussed in ‘Control Layer’ unit), the storage virtualization software can perform tasks such as virtual volume creation apart from creating virtual arrays. This software provide a single control point to the entire storage infrastructure, enabling automated and policy based management. The storage virtualization can also be available as hypervisor’s capability, which enables creating virtual disk that appears to the operating systems as physical disk drives.
• #22: A virtual machine (VM) is a logical compute system that, like a physical compute system, runs an operating system (OS) and applications. A VM is created by a hosted or a bare-metal hypervisor installed on a physical compute system. A VM has a self-contained operating environment, comprising OS, applications, and virtual hardware, such as a virtual processor, memory, storage, and network resources. An OS, called—a ‘guest’ OS—is installed on a VM in the same way like it is installed on a physical compute system. From the perspective of the guest OS, the VM appears as a physical compute system. As discussed in lesson 1, a virtual machine monitor (VMM) is responsible for the execution of a VM. Each VM has a dedicated VMM. Each VM has its own configuration for hardware, software, network, security, and so on. The VM behaves like a physical compute system, but does not have direct access either to the underlying host OS (when a hosted hypervisor is used) or to the hardware of the physical compute system on which it is created. The hypervisor translates the VM’s resource requests and maps the virtual hardware of the VM to the hardware of the physical compute system. For example, a VM’s I/O requests to a virtual disk drive are translated by the hypervisor and mapped to a file on the physical compute system’s disk drive. Compute virtualization software enables creating and managing several VMs—each with a different OS of its own—on a physical compute system or on a compute cluster. In a cloud environment, a provider typically provisions VMs to consumers to deploy their applications. The VM hardware and software are configured to meet the application’s requirements. The VMs of consumers are isolated from each other so that the applications and the services running on one VM do not interfere with those running on other VMs. The isolation also provides fault tolerance so that if one VM crashes, the other VMs remain unaffected.
• #23: The slide lists the common types of virtual networks.
• #24: A virtual LAN (VLAN) is a virtual network consisting of virtual and/or physical switches, which divides a LAN into smaller logical segments. A VLAN groups the nodes with a common set of functional requirements, independent of the physical location of the nodes. In a multi-tenant cloud environment, the provider typically creates and assigns a separate VLAN to each consumer. This provides a private network and IP address space to a consumer, and ensures isolation from the network traffic of other consumers. Traditionally in a physical network, a router is typically used to create a LAN and the LAN is further segmented by using switches and hubs. In a physical LAN, the nodes, switches, and routers are physically connected to each other and must be located in the same area. VLANs enable a network administrator to logically segment a LAN, and the nodes do not have to be physically located on the same LAN. For example, a cloud provider may place the VMs of a consumer in the same VLAN, and the VMs may be on the same compute system or different ones. Also, if a node is moved to another location, depending on the VLAN configuration, it may still stay on the same VLAN without requiring any reconfiguration. This simplifies network configuration and administration. A node (VM, physical compute system, or storage system) may be a member of multiple VLANs, provided the OS, hypervisor, and storage array OS support such configurations. To configure VLANs, an administrator first defines the VLANs on the physical and virtual switches. Each VLAN is identified by a unique 12-bit VLAN ID (as per IEEE specification 802.1Q). The next step is to configure the VLAN membership based on different techniques, such as port-based, MAC-based, protocol-based, IP subnet address-based, or application-based. In the port-based technique, membership in a VLAN is defined by assigning a VLAN ID to a physical or virtual switch port or port group. In the MAC-based technique, the membership in a VLAN is defined on the basis of the MAC address of the node. In the protocol-based technique, different VLANs are assigned to different protocols based on the protocol type field found in the OSI Layer 2 header. In the IP subnet address-based technique, membership is based on the network IP subnet address of the OSI Layer 3 header. In the application-based technique, a specific application, for example, a file transfer protocol (FTP) application can be configured to execute on one VLAN. A detailed discussion on these VLAN configuration techniques is beyond the scope of this course.
• #25: A private VLAN (PVLAN) is an extension of the VLAN standard and further segregates the nodes within a VLAN into sub-VLANs. A PVLAN is made up of a primary VLAN and one or more secondary (or private) VLANs. The primary VLAN is the original VLAN that is being segregated into smaller groups. Each secondary PVLAN exists only inside the primary VLAN. It has a unique VLAN ID and isolates the OSI Layer 2 traffic from the other PVLANs. Primary VLANs are promiscuous, which means that ports on the PVLANs can communicate with ports configured as the primary VLAN. Routers are typically attached to promiscuous ports. There are two types of secondary PVLANs within a primary VLAN: Isolated and Community. Isolated: A node attached to a port in an isolated secondary PVLAN can only communicate with the promiscuous PVLAN. Community: A node attached to a port in a community secondary PVLAN can communicate with the other ports in the same community PVLAN as well as with the promiscuous PVLAN. Nodes in different community PVLANs cannot communicate with each other. To configure PVLANs, the PVLAN feature must be supported and enabled on a physical switch or a distributed virtual switch. To create PVLANs, the administrator first creates standard VLANs on a switch, and then configures the VLANs as primary and secondary. The figure on the slide illustrates how different types of PVLANs enable and restrict communications between VMs (nodes) that are connected to a distributed virtual switch. PVLANs enable a cloud provider to support a larger number of consumers and addresses the issues with scalability encountered in VLANs. If a service provider assigns one VLAN per customer, it limits the number of consumers that can be supported. Also, a block of addresses are assigned to each consumer VLAN, which can result in unused IP addresses. Additionally, if the number of nodes in the VLAN increases, the number of assigned addresses may not be large enough to accommodate them. In a PVLAN, all members share a common address space, which is allocated to the primary VLAN. When nodes are connected to secondary VLANs, they are assigned IP addresses from the block of addresses allocated to the primary VLAN. When new nodes are added in different secondary VLANs, they are assigned subsequent IP addresses from the pool of addresses. (Cont'd)
• #26: PVLANs also provide an additional layer of security between the nodes that belong to the same VLAN. They also simplify the administrative overhead of maintaining access control lists (ACLs) on different VLANs, which restrict network traffic movement between VLANs. Further, a provider can ensure network traffic isolation by configuring isolated PVLANs. Without isolated PVLANs, segregating network traffic from different consumers requires creating a large number of discrete VLANs.
• #27: A stretched VLAN is a VLAN that spans across multiple sites over a WAN connection. In a typical multi-site environment, two sites are connected over an OSI Layer 3 WAN connection and all network traffic between them is routed. Because of the routing, it is not possible to transmit OSI Layer 2 WAN traffic between the nodes in the two sites. A stretched VLAN extends a VLAN across the sites and enables nodes in the two different sites to communicate over a WAN as if they are connected to the same network. Stretched VLANs also allow the movement of VMs between sites without having to change their network configurations. This enables the creation of high-availability clusters, VM migration, and application and workload mobility across sites. For example, in the event of a disaster or during the maintenance of one site, a provider typically moves VMs to an alternate site. Without a stretched VLAN, the IP addresses of the VMs must be changed to match the addressing scheme at the other site. Stretched VLANs may be created by simply connecting two sites using long distance fiber and can be configured using different methods depending upon the underlying WAN technology. Stretched VLANs may be created by simply connecting two sites using long distance fiber, Dense Wave Division Multiplexing (DWDM), Coarse Wave Division Multiplexing (CWDM), multi-protocol label switching (MPLS) network, and IP network. An elaboration of these methods is beyond the scope of this course.
• #28: A VXLAN is a OSI Layer 2 overlay network built on a OSI Layer 3 network. An overlay network is a virtual network that is built on top of existing network. VXLANs, unlike stretched VLANs, are based on LAN technology. VXLANs use the MAC Address-in-User Datagram Protocol (MAC-in-UDP) encapsulation technique. In this scheme, a VXLAN header is added to the original Layer 2 (MAC) frame, which is then placed in a UDP-IP packet and tunneled over a Layer 3 network. Communication is established between two tunnel end points called Virtual Tunnel Endpoints (VTEPs). At the transmitting node, a VTEP encapsulates the network traffic into a VXLAN header and at the destination node, a VTEP removes the encapsulation before presenting the original Layer 2 packet to the node. VXLANs enable the creation of a logical network of nodes across different networks. In case of VM communication, the VTEP is built into the hypervisor on the compute system hosting the VMs. VXLANs enable the separation of nodes, such as VMs, from physical networks. They allow the VMs to communicate with each other using the transparent overlay scheme over physical networks that could span Layer 3 boundaries. This provides a means to extend a Layer 2 network across sites. The VMs are unaware of the physical network constraints and only see the virtual Layer 2 adjacency. Nodes are identified uniquely by the combination of their MAC addresses and a VXLAN ID. VXLANs use a 24-bit VXLAN ID, which makes it theoretically possible to have up to 16 million Layer 2 VXLANs co-existing on a common Layer 3 infrastructure. VXLANs make it easier for administrators to scale a cloud infrastructure while logically isolating the applications and resources of multiple consumers from each other. VXLANs also enable VM migration across sites and over long distances.
• #29: A virtual SAN (VSAN) or virtual fabric is a logical fabric created on a physical FC or FCoE SAN. A VSAN enables communication between a group of nodes with a common set of requirements, independent of their physical location in the fabric. A VSAN functions conceptually in the same way as a VLAN. Each VSAN behaves and is managed as an independent fabric. Each VSAN has its own fabric services, configuration, and set of FC addresses. Fabric-related configurations in one VSAN do not affect the traffic in another VSAN. Also, the events causing traffic disruptions in one VSAN are contained within that VSAN and are not propagated to the other VSANs. Similar to a stretched VLAN, a VSAN may be extended across sites by using long distance fiber, DWDM, CWDM, or FCIP links to carry the FC frames.
• #30: To configure VSANs on a fabric switch, the VSANs are first defined with specific VSAN IDs. Then the F_Ports on the switch are assigned the VSAN IDs to include them in the respective VSANs. If an N_Port connects to an F_Port that belongs to a VSAN, it becomes a member of that VSAN. Note: VSAN vs. Zone Both VSANs and zones enable node ports within a fabric to be logically segmented into groups. But they are not same and their purposes are different. There is a hierarchical relationship between them. An administrator first assigns physical ports to VSANs and then configures independent zones for each VSAN. A VSAN has its own independent fabric services, but the fabric services are not available on a per-zone basis.
• #31: The FCoE protocol enables transmission of FC SAN traffic through a LAN that supports Data Center Bridging (DCB) functionalities. The FC frames remain encapsulated into Ethernet frames during transmission through the LAN. If VLANs and VSANs are created on the LAN and FC SAN respectively, a mapping is required between the VLANs and VSANs. The mapping determines which VLAN will carry FC traffic that belongs to a VSAN. The mapping of VSAN to VLAN is performed at the FCoE switch. Multiple VSANs are not allowed to share a VLAN. Hence, a dedicated VLAN must be configured at the FCoE switch for each VSAN. Also, it is recommended that VLANs that carry regular LAN traffic should not be used for VSAN traffic. The figure on the slide shows an example of a mapping between VLANs and VSANs. In the example, the FCoE switch is configured with four VLANs: VLAN 100, VLAN 200, VLAN 300, and VLAN 400. The Ethernet switch is configured with two VLANs: VLAN 100 and VLAN 200. Both VLAN 100 and VLAN 200 transfer regular Ethernet traffic to enable compute-to-compute communication. The fabric switch has VSAN 100 and VSAN 200 configured. To allow data transfer between the compute system and the FC fabric through the FCoE switch, VSAN 100 and VSAN 200 must be mapped to VLANs configured on the FCoE switch. Since VLAN 100 and VLAN 200 are already being used for LAN traffic, VSAN 100 and VSAN 200 should be mapped to VLAN 300 and VLAN 400, respectively.
• #41: According to Organization for the Advancement of Structured Information Standards (OASIS), a reference model is an abstract framework for understanding the significant relationships among the entities of some environment, and for the development of consistent standards or specifications supporting that environment. A reference model is based on a small number of unifying concepts and may be used as a basis for education and explaining standards. A reference model is not directly tied to any standards, technologies, or other concrete implementation details, but it does seek to provide a common semantics that can be used unambiguously across and between different implementations. Key goals of reference model are: Conveys fundamental principles and basic functionality of a system it represents Facilitates efficient communication of system details between stakeholders Provides a point of reference for system designers to extract system specifications Enhances an individual’s understanding of the representative system Documents the system for future reference and provides a means for collaboration
• #42: The cloud computing reference model is an abstract model that characterizes and standardizes the functions of a cloud computing environment by partitioning it into abstraction layers and cross-layer functions. This reference model groups the cloud computing functions and activities into five logical layers and three cross-layer functions. The five layers are physical layer, virtual layer, control layer, service orchestration layer, and service layer. Each of these layers specifies various types of entities that may exist in a cloud computing environment, such as compute systems, network devices, storage devices, virtualization software, security mechanisms, control software, orchestration software, management software, and so on. It also describes the relationships among these entities. The three cross-layer functions are business continuity, security, and service management. Business continuity and security functions specify various activities, tasks, and processes that are required to offer reliable and secure cloud services to the consumers. Service management function specifies various activities, tasks, and processes that enable the administrations of the cloud infrastructure and services to meet the provider’s business requirements and consumer’s expectations.