Chapter 2.
Chapter 2.
Migrating into a Cloud
Migrating into a Cloud
(a)
(a)
Introduction
Introduction
- Cloud computing: a disruptive model of IT
* Part technology and part business model
* Raised the IT expectations of small and medium enterprises * Large companies are deeply debating it
* Questions:
# When and how to migrate one’s application into a cloud?
# What part or component of the IT application to migrate into a cloud and what not to migrate into a cloud
# What kind of customers really benefit from migrating their IT into the cloud
- Promise of cloud computing services
* Cloudonomics:
# Pay per use: lower cost barrier # On demand resources: autoscaling
# CAPEX vs OPEX: no capital expenses (CAPEX) and only operational expenses (OPEX) # SLA driven operations: much lower TCO (Total cost of ownership)
# Attractive NFR (Not-for-resale) support: availability, reliability * Technology:
# Infinite elastic availability: compute/storage/bandwidth # Automatic usage monitoring and metering
# Jobs/tasks virtualized and transparently movable
# Integration and interoperability support for hybrid operations # Transparently encapsulated and abstracted IT features
- Features of cloud services
* Typically web-oriented
* Represents seasonal IT demands * Non-mission critical
* Low demand of security
(b)
(b)
Broad approaches to migrating into the cloud
Broad approaches to migrating into the cloud
- Migrating to cloud is a trend
* Top 10 strategies to watch for 2013 * Questions:
# Cost to migrate to cloud appropriate?
- Capital expense eliminated, operational expenses incurred only # Total cost of ownership of data reduced?
- Keep data on cloud provider’s or the company’s data center
- Why migrate?
# An independent application on the cloud
# Code of the application needs to be modified to be on the cloud # The application needs to be redesigned
# The architecture of the application needs to be redeveloped
# The application stays as it is, but the usage needs to be changed * In summary:
# Migration can happen in one of the five levels: application, code, design, architecture, and usage
# Representation of the migration of the enterprise application: P → P'C + P'l → P'OFC + P'l
- P: application before migration
- P'C: part of the application migrated into the cloud
- P'l: part of the application staying local
- P'OFC: the application part optimized for cloud
- If the entire application is migrated, P'l: is null
* Cost-effective analysis: analyze the migration representation of the application, e.g.: # IaaS migration: 30 use-case scenarios
# PaaS migration: 20 use-case scenarios # SaaS migration: only usage
→ What’s the total cost? * Cloudonomics
→ Cost cutting in CAPES and OPEX
# Cost of migration: is it economically feasible or tenable? # License fee: SLA compliance
# Price of the cloud offerings
- Deciding on the cloud migration
* Design a questionnaire with several classes of key questions that impact the IT due to the migration
* Pose the questionnaire to selected people (technical and business experts) * Modeling the decision making:
# Suppose:
- There are M classes of questions, Mi, i = 1..M
- Class Mi has Ni questions
- Each class is assigned a certain relative weight Bi, i = 1..M
# The decision making can be modeled as an M×N weight matrix:
Cl ≤
∑
i=1 M Bi(
∑
j=1 Ni Aij Xij)
≤ Ch - E.g., B1(A11X11+A12X12+…) + B2(A21X21+A22X22+A23X23+…) + … - Cl: lower weight threshold- Ch: higher weight threshold
- Xij : the degree to which that answer to the question is relevant and applicable, 0 ≤ Xij ≤ 1
(c)
(c) The seven-step model of migration into a cloud
The seven-step model of migration into a cloud
- Seven-step model of migration
* Conduct cloud migration assessment
# Understand the migration issues at the application level or the code, the design, the architecture, or usage levels
# Cost of migration: ROI (Return of investment), TCO, … * Isolate the dependencies
# Isolate all systematic and environmental dependencies of the enterprise application components within the captive data center
* Map the messaging and environment
# Message map: displaying detailed, hierarchically organized responses to anticipated questions or concerns
# Generate the mapping constructs between what shall remain in the local captive data center and what goes onto the cloud
* Re-architect and implement the lost functionalities
# Perhaps some functionality may be lost due to migration
# Some part of the enterprise application may need to be re-architect, redesigned, and reimplemented on the cloud
* Leverage cloud function and features
# Leverage the intrinsic features of the cloud computing service to augment the
enterprise application * Test the migration
# Test the new form of the enterprise
application (both on the captive data center and on the cloud as well)
* Iterate and optimize
# Iterate and optimize the process
- Migration risks and mitigation
* In the seven-step model:
# Test step: identify the key migration risks # Optimization step: mitigate the migration
risks
* Two categories of risk: # General migration risks
- Performance monitoring and tuning
- Business continuity and disaster recovery of the cloud computing service - Compliance with standards and governance issues
- QoS parameters and SLA
- Ownership, transfer, and storage of data in the application
- Portability and interoperability issues that would mitigate potential vendor lock-ins # Security-related migration risks
- Trust and privacy
- The right to obtain execution logs and audit trails as a detailed level - Problems due to multi-tenancy