Migrating cloud resources

1. A computer implemented method of cloud migration, the method comprising:

accessing repository data fragments distributed across one or both of a first cloud and a network;

combining the repository data fragments into repository data;

discovering first resources of the first cloud, and a dependency between the first resources, and second resources of a second cloud;

generating a model of the first cloud based on the discovered first resources of the first cloud and the repository data;

discovering second resources of a second cloud;

generating a migration map between the first cloud and the second cloud based on the discovered first and second resources and based on the repository data;

analyzing the migration map for any errors and to determine the validity of the migration map;

recursively modifying the migration map in response to determining that the migration map is invalid or contains an error, wherein the recursive modification comprises a manual input, an additional automated resource discovery, or a combination thereof; and

migrating the first resources to the second cloud based on the modified migration map.

2. The computer implemented method of claim 1 wherein the repository data represents a migration archetype comprising an archetypical migration map or an archetypal cloud topology.

3. The computer implemented method of claim 1 wherein the repository data represents resources and dependencies previously discovered in a past migration to the first cloud.

4. The computer implemented method of claim 1 wherein the data fragments are distributed across the first cloud.

5. The computer implemented method of claim 1 wherein the data fragments are distributed across the network, the network being the Internet.

6. The computer implemented method of claim 1 wherein the repository data is represented in an ontology language.

7. The computer implemented method of claim 6 wherein the ontology language is OWL (Web Ontology Language) or RDF (Resource Description Framework).

8. The computer implemented method of claim 6 wherein the repository data fragments are accessed using SPARQL (SPARQL Protocol and RDF Query Language).

9. The computer implemented method of claim 1 further comprising generating a model of the first cloud based on the discovered first resources and first dependencies, the model expressed in declarative language.

10. The computer implemented method of claim 1 further comprising discovering a capacity attribute of the second cloud, wherein the migration map is generated based on the discovered capacity attribute.

11. The computer implemented method of claim 1 wherein the discovery is automatic.

12. A non-transitory computer readable storage medium including executable instructions that, when executed by a processor, cause the processor to:

access repository data fragments distributed across at least one of a first cloud and a network;

combine the repository data fragments into repository data;

discover a first cloud topology of the first cloud, the first cloud topology representing resources of the first cloud and a dependency between the first resources;

generate a model of the first cloud based on the discovered first cloud topology and the repository data;

discover a second cloud topology of a second cloud;

determine and generate a migration map between the first and second clouds based on the first and second cloud topologies and the repository data;

analyze the migration map for any errors and to determine the validity of the migration map;

recursively modify the migration map in response to determining that the migration map is invalid or contains an error, wherein the recursive modification comprises a manual input, an additional automated resource discovery, or a combination thereof; and

migrate the resources to the second cloud based on the modified migration map.

13. The non-transitory computer readable storage medium of claim 12 wherein the repository data represents at least one of an archetypical migration map, an archetypal cloud topology, and a resources and dependencies previously discovered in a past migration to the first cloud.

14. A cloud controller comprising:

a memory; and

a processor coupled to the memory to:

access data fragments distributed across one or both of a first cloud and a network;

combine the data fragments into repository data;

discover first resources of the first cloud and and a dependency between the first resources;

generate a model of the first cloud based on the discovered first resources of the first cloud and the repository data;

discover second resources of a second cloud and a capacity attribute of a second cloud;

generate a migration map between the first cloud and the second cloud based on the discovered first resources, the discovered capacity attribute, and the repository data;

analyze the migration map for any errors and to determine the validity of the migration map;

recursively modify the migration map to increase accuracy of the migration map in response to determining that the migration map is invalid or contains an error, wherein the recursive modification comprises a manual input, an additional automated resource discovery, or a combination thereof; and

migrate the first resources to the second cloud based on the modified migration map.

15. The cloud controller of claim 14 wherein the migration is based on the modified migration map.