ATMOSPHERE Webinar: Combining Clouds with Fogbow (Francisco Brasileiro)
- 1. Co-funded by the European Commission Horizon 2020 - Grant #777154
Brazilian Ministry of Science, Technology & Innovation
Coordinated calls between Europe and Brazil
Webinar series
Combining clouds with
Fogbow
atmosphere-eubrazil.eu @AtmosphereEUBR
- 2. Speakers
• Francisco Brasileiro
– Professor at Universidade Federal de Campina
Grande in Brazil
– Brazilian coordinator of the ATMOSPHERE project
• Leandro Ciuffo
– Associate Director of Advanced Internet in the
Directorate of R&D at RNP (Brazil's National
Research and Education Network)
2
- 3. Co-funded by the European Commission Horizon 2020 - Grant #777154
Brazilian Ministry of Science, Technology & Innovation
Coordinated calls between Europe and Brazil
The Fogbow Middleware
Francisco Brasileiro
fubica@dsc.ufcg.edu.br
atmosphere-eubrazil.eu @AtmosphereEUBR
- 4. Fogbow in a nutshell
• Fogbow is a microservice-based midleware
that supports multiple ways to combine
clouds, including:
– Cloud brokering
– Hybrid clouds
– Federated clouds
• Open source
• Completely customizable via plugins
4
- 5. IaaS deployment models
• Public cloud providers
– Shared multi-site infrastructures
– Economy of scale leads to very efficient providers
– Provides higher elasticity to users
– Pay-as-you-go pricing scheme reduces upfront
investments to users
– Might raise security and privacy concerns
5
- 6. IaaS deployment models
• Private cloud providers
– Dedicated (typically) single-site infrastructures
– Brings no new security and privacy concerns
– Smaller than public providers, thus tend to be
less efficient
– Normally, less elastic than public providers
• More restrictive usage quotas
6
- 7. Combined deployment models
• Multi-clouds (brokers)
– Allow multiple public clouds to be used,
increasing the portfolio of services available to
users
– Allow resource allocation driven by some
objective function (eg. minimize cost)
– If implemented in-house, may mitigate vendor
lock-in issue of public clouds
7
- 8. Combined deployment models
• Hybrid clouds
– Typically, extends the infrastructure of a private cloud
by allowing some of the workload to be outsourced
to a public cloud
– Improves the efficiency of the private cloud
– Provides access to multiple geographically distributed
sites
• Useful for fault tolerance and attend geo-sensitive
workloads
– Sensitive workload might need to run in the private
part of the infrastructure
8
- 9. Combined deployment models
• Federation/community of private clouds
– Shared multi-site infrastructures
• Fine control on how sharing is performed at each
autonomous site (from the providers point of view)
– Depending on how members trust each other,
may not raise extra security and privacy concerns
– May allow for higher elasticity
– More flexibility in defining business models
9
- 10. Combining clouds: cloud brokering
10
AWS datacenter at Sao Paulo
(runs AWS services)
RNP Compute Service
(runs CloudStack)
- 17. Visit website for more information
17
Check it
out here!
Get on
board!
Download and use it!