Proven Cloud Architecture and Architectural Layers
From our work in helping organizations whose mission it is to deliver successful production Eucalyptus clouds, we have documented a catalog of private cloud reference architecture and architectural layers that can be used as the foundation for production deployments.
Cloud architecture and architectural layers are categorized by use case and include:
- Description of the intended use of the resulting production deployment
- Discussion on the balance between virtual machines, storage, and certain features
- Discussion of all technologies that are required to successfully run a production Eucalyptus cloud
Not all deployments will exactly implement one of the private cloud reference architecture deployments, but all production deployments should start off with one of the architectures below.
Reference Architectures for Eucalyptus 4
Eucalyptus 4 includes new features and architectural changes intended to support larger scale clouds. New reference architectures being developed will help users incorporate features like Edge Networking and the Object Storage Gateway into their deployment plans.
View the following Eucalyptus 4 Reference Architectures:
- Eucalyptus 4 Reference Architecture DRAFTS.
- Eucalyptus 4 Reference Architecture Dev/Test Small on HP ProLiant
- Eucalyptus 4 Reference Architecture Dev/Test Small Detailed on HP ProLiant
- Eucalyptus 4 Reference Architecture Dev/Test Large on HP ProLiant
Not all features of Eucalyptus are exercised by these private cloud reference architectures. Many features of Eucalyptus (multi-cluster tunnelling, LDAP integration) are present for integration purposes, but are not implemented as 'deployment best practice' if they are not needed.
Each of the cloud architectures and their architectural layers specify 'maximums' that the deployment can support. If the maximums in place for each private cloud reference architecture need to be substantially exceeded for a deployment, the deployment should be taken into consideration as one that will require devoted, customized attention to ensure its production readiness.