This document presents a framework for intelligently placing datacenters for internet services. It discusses parameters like costs, response time, and emissions that are considered. The framework formulates the placement problem by taking inputs like user numbers, servers, and existing datacenters. It evaluates solutions like linear programming and simulated annealing to find an optimal placement configuration with minimum cost. A placement tool is developed that considers location-dependent data. The tool is used to evaluate placements and tradeoffs around latency, availability, consistency, emissions and energy efficiency. The document concludes that the framework and tool can automatically place datacenters by optimizing multiple objectives and parameters.
1. EEDC
Execution
34330
Intelligent Placement of
Datacenters for Internet
Environments for Services
Distributed
Computing
European Master In Distributed
Computing (EMDC)
Homework number: 6
Paper Presentation, EEDC
Ioanna Tsalouchidou –
ioannatsalouchidou@gmail.com
2. Contents
• Datacenters
• The Problem
• Framework
• The placement tool
• Evaluation
• Tradeoffs
• Conclusions
2
3. Datacenters
• Where Google, Yahoo, Microsoft etc host services
• Geographically distributed
• Enormous cost of provisioning
• Locations selected intelligently
3
4. The Problem
Selecting the location
●
●
Optimization problem
●
Solutions
Efficiency and accuracy
●
Characterization of the areas
●
Quantification and tradeoffs
●
4
6. Framework
Formulating the problem:
●
●
Input
●
Max number of servers
●
Utilization of servers
●
Number of users
●
Redundancy
●
Network latency
●
Output
●
Optimal cost
●
Max number of servers/location
●
Number of servers/population center
●
Existing datacenters
6
7. Framework
Solution approaches
●
●
Simple linear programming, LP0
●
Pre-Set linear programming, LP1
●
Brute force
●
Heuristic based on LP
●
Simulated annealing plus, SA+LP1
●
Optimized SA+LP1, OSA+LP1
7
8. The Tool
Location Dependent Data
●
●
Network backbones
●
Power plants, transmission lines, CO2 emissions
●
Electricity, land, water, temperature
●
Missing data
8
9. The Tool
Characteristics
●
●
Datacenter size, cooling, PUEs
●
Connection costs
●
Building costs
●
Land costs
●
Water costs
●
Servers, internal networking costs
●
Staff costs
9
14. Conclusion
Automatic placement of Datacenters
●
Optimization framework
●
Solution approaches
●
Tradeoffs
●
Future: more fine grained input data automatically
●
14
15. EEDC
Execution
34330
Intelligent Placement of
Datacenters for Internet
Environments for Services
Distributed
Computing
European Master In Distributed
Computing (EMDC)
Homework number: 6
Paper Presentation, EEDC
Ioanna Tsalouchidou –
ioannatsalouchidou@gmail.com