Federation » History » Version 3
Tom Clegg, 06/22/2018 08:57 PM
| 1 | 1 | Tom Clegg | h1. Federation |
|---|---|---|---|
| 2 | |||
| 3 | In a federation, clusters cooperate with one another to enable distributed workflows that are coordinated by a remote client, without having to funnel data/code/notifications through the remote client. |
||
| 4 | |||
| 5 | Canonical example (draft!): |
||
| 6 | * The user agent has a slow network connection to several clusters which, although they have reasonably fast network connections between them, must not have unrestricted access to one another's data. |
||
| 7 | * The user must run an analysis A on Cluster A. The inputs to A are large datasets, and they are the outputs of analyses B and C, which must run on Cluster B and C respectively. |
||
| 8 | |||
| 9 | Subtopics and related features: |
||
| 10 | * [[Federation use cases]] |
||
| 11 | 2 | Tom Clegg | * [[Federation implementation roadmap]] |
| 12 | 1 | Tom Clegg | * [[Federated identity]] |
| 13 | * [[Distributed workflows]] |
||
| 14 | * [[Routing multi cluster requests]] |
||
| 15 | 3 | Tom Clegg | * [[Authentication for a multi-cluster workflow]] |