Review of Cloud Computer (EE4221)

big_data.md

Definition

Volume, Variety, and Velocity

Unstructed

Analytic life cycle

1. Discovery
2. Data preparation
3. Extract the data
4. Data cleaning and preprocessing
5. Make sure you have enough good quality data
6. Study your data
7. Model Planning
8. Determine the methods, techniques, and workflow you intend to follow
9. Determine the variables and models you are likely to use
10. Model Building
11. Develop data sets for testing, training and production
12. Setup the environment for executing the the models and workflows
13. Communicate results
14. Determine if you succeeded or failed based on the criteria you defined in Discovery phase
15. Identify the key findings and quantify the business value of the results
16. Develop a summary and communicate it with your stakeholders
17. Operationalize
18. Deliver final reports, briefings, code, and technical documents
19. Run a pilot project, and implement a production environment
20. Demonstare the clear value
21. Translate the results into a language that speaks to your audience - Visualization

Challenges with unstructured data

• Large amount of data
• Evolving and distributed data structures
• Overhead to ensure data integrity and schema consistency
• Scalability

public_cloud.md

Cloud Computing Characteristics

• on-demand self-service
• broad network access
• resource pooling
• rapid elasticity
• measured service

Cloud Strategies

Public cloud

resources are dynamically provisioned on a fine-grained, self-service basis over the Internet

Community cloud

pooled resource cloud that combines the resources of multiple community users

Private cloud

• owned by comporations which uses them internally
• may be completely behind the firewall so that no external access is possible

Hybrid cloud

• consists of multiple internal and/or external providers
• a combination of public and private clouds: utilizes public cloud for non-critical information handling and private cloud for dealing critical information
• public cloud is also used to balance workload (high peaks are run on public cloud)

Level

IaaS

• Scalability
• Error Recovery
• Time Back
  • With the infrastructure being hosted offsite, you can focus on value-added tasks such as planning and development.
• Efficient payment model

Paas

• Jump start development
• No maintenance cost
• Lower risk factor

How to select

• Ensure that it can be configured to meet the changeing needs
• Cloud vendor should provide high SLA guarantees for their services
• Learn about the cloud provider’s key risks and performance indicators
• Learn how easily you can switch to another provider if needed
• Check the quality of customer service and support

virtual_compute.md

Benefit

• Server consolidation
  • Multiple virtual machines can be run in the same physical servers
• Isolation
  • Virtual machines remain completely isolated from each other
• Encapsulation
  • A VM can be moved and copied from one location to the another just like a file
• Hardware independence
  • A VM is configured with virtual components
• Reduced cost
  • Compute virtualization reduces direct costs

Techniques

Full Virtualization

• Virtual Machine Monitor runs in privileged Ring 0
• Guest OS is not aware of being virtualized

Paravirtualization

• Guest operating system (OS) knows that it is virtualized
• Guest OS runs in Ring 0

Hardware Assisted Virtualization

• Achieved by using hypervisor-aware CPU to handle privileged instructions
  • Reduces virtualization overhead caused due to full and paravirtualization

virtual_network.md

VLAN

VLAN

Facts

• VLAN traffic is transferred through routers
• A computer may be members of multiple VLANs

Benefit

• Controls broadcast activity and improves network performance
• Simplifies management
• Increases security levels
• Provides higher utilization of switch and reduces CAPEX

VLAN Trunking

Define

It is a technology that allows traffic from multiple VLANs to traverse a single network connection.

Facts

• Single connection & Single port
• Trunk port is included to all VLANs
• Enabled by tagging

Benefit

• Eliminates the need for dedicated network link(s) for each VLAN
• Reduces inter-device links when the devices have more than one VLAN

VLAN Tagging

Define

It is a process of inserting or removing a marker (tag) with VLAN- specific information (VLAN ID) into the Ethernet frame

VLAN vs Nextork Virtualization

VLAN is used in both physical networks (before network virtualization) and virtualized networks (after network virtualization)

Network Virtualization

VM Network

Facts

• Resides inside physical server
• Consists of logical switches called “virtual switches”
• Provides connectivity among VMs inside a physical server
• Provides connectivity to Hypervisor kernel
• Connects to physical network
• VM and physical networks are virtualized to create virtual networks; for example: virtual LAN, virtual SAN

Tools

• Physical switch Operation System
• Hypervisor

Benefit

• Enhance security
  • Restricts access to nodes in a virtual network from another virtual network
  • Isolates sensitive data from one virtual network to another
• Enhance performance
  • Restricts network broadcast and improves virtual network performance
• Improves manageability
  • Allows configuring virtual networks from a centralized management workstation using management software
  • Eases grouping and regrouping of nodes
• Improves utilization and reduces Capital Expenditure (CAPEX)
  • Enables multiple virtual networks to share the same physical network, which improves utilization of network resource
  • Reduces the requirement to set up separate physical networks for different node groups

Component

Types

• V NIC: VMs to VM Network (V switch)
• V HBA: VM to disk/LUN
• V switch
• Physical ..

Facts

• Physical NIC has no IP address or MAC address
• Physical NIC does not modify the frames
• V NIC has IP address & MAC address
• 
• Physical NICs are not shared between virtual switches
• No direct connection between virtual switches
• Frames may be transferred between virtual switches via a VM

Ports and Port Group

Facts

• VMs connected to a VM port group share common configuration

Distributed Virtual Switch

Benefit

• Centralizes VM network management
• Maintains network policies during VM migration

Tagging

External Switch Tagging (EST)

• external switch handles adding and removing the VLAN tags
• virtual switch will handle untagged frames only

Virtual Switch Tagging (VST)

• virtual switch itself is inspecting and adding/removing the VLAN tags
• physical uplink port (that the virtual switch connected to) must be configured as a trunk port