Dynamic application consistent data restoration
Restore operations in containerized environments are disclosed. An ephemeral instance of an application is created and a datastore is mounted to the ephemeral instance. The ephemeral instance is not accessible to users or application. The backup data is restored to the datastore. Once restored, the datastore is then mounted to a production instance and production resumes.
1. A method for restoring backup data in a container runtime environment, the method comprising:
determining an application type of an application associated with backup data based on artifacts obtained from an artifact storage;
creating an ephemeral application instance of the application based on the application type and the artifacts;
mounting a data store to an ephemeral application instance;
restoring the backup data to the data store by the ephemeral application instance, wherein an entry point to the data store is isolated from user access;
running a production instance of the application in the container runtime environment; and
mounting the datastore to the production instance for operation.
2. The method of claim 1 , wherein the artifacts include one or more of declarative models of an application state, application manifests, and/or storage requirements.
3. The method of claim 2 , wherein the declarative models specify a desired result rather than steps to achieve the desired result or wherein the declarative models specify a desired state rather than steps to achieve the desired state.
4. The method of claim 1 , wherein the manifest includes a container name and type, volumes and their respective mappings, environment variables, port mappings, and/or registry values or wherein the manifest includes information describing a deployment and relationship of the data to an application specific configuration.
5. The method of claim 1 , wherein the datastore is not accessible to end-users or other applications when mounted to the ephemeral application instance.
6. The method of claim 1 , further comprising dynamically generating a manifest based on the artifacts.
7. The method of claim 1 , further comprising unmounting the data store from the ephemeral application instance after the backup data has been restored to the data store.
8. The method of claim 1 , wherein the artifacts are configured such that the backup data can be restored without human interaction and knowledge of the application type.
9. The method of claim 8 , wherein the artifacts allow the backup data to be restored without allowing access to the data to be restored until the data is fully prepared and in a consistent state.
10. The method of claim 1 , where determining an application type includes inferring the application type based on the artifacts.
11. A non-transitory storage medium having stored therein instructions that are executable by one or more hardware processors to perform operations comprising:
determining an application type of an application associated with backup data based on artifacts obtained from an artifact storage;
creating an ephemeral application instance of the application based on the application type and the artifacts;
mounting a data store to an ephemeral application instance;
restoring the backup data to the data store by the ephemeral application instance, wherein an entry point to the data store is isolated from user access;
running a production instance of the application in the container runtime environment; and
mounting the datastore to the production instance for operation.
12. The non-transitory storage medium of claim 11 , wherein the artifacts include one or more of declarative models of an application state, application manifests, and/or storage requirements.
13. The non-transitory storage medium of claim 12 , wherein the declarative models specify a desired result rather than steps to achieve the desired result or wherein the declarative models specify a desired state rather than steps to achieve the desired state.
14. The non-transitory storage medium of claim 11 , wherein the manifest includes a container name and type, volumes and their respective mappings, environment variables, port mappings, and/or registry values or wherein the manifest includes information describing a deployment and relationship of the data to an application specific configuration.
15. The non-transitory storage medium of claim 11 , wherein the datastore is not accessible to end-users or other applications when mounted to the ephemeral application instance.
16. The non-transitory storage medium of claim 11 , further comprising dynamically generating a manifest based on the artifacts.
17. The non-transitory storage medium of claim 11 , further comprising unmounting the data store from the ephemeral application instance after the backup data has been restored to the data store.
18. The non-transitory storage medium of claim 11 , wherein the artifacts are configured such that the backup data can be restored without human interaction and knowledge of the application type.
19. The non-transitory storage medium of claim 18 , wherein the artifacts allow the backup data to be restored without allowing access to the data to be restored until the data is fully prepared and in a consistent state.
20. The non-transitory storage medium of claim 11 , where determining an application type includes inferring the application type based on the artifacts.