System and method for storage unavailability tolerant backup
Systems, devices, and methods for reducing the impact of temporary unavailability of storages in a backup storage system. The system may store different types of backups depending on the availability storages. The first type of backup, which may rely on a previous backup stored in a first storage that is available for restoration purposes, may be stored on the first storage. When the first storage becomes temporarily unavailable, a second type of backup may be stored in a second storage that is available that requires a previous backup stored in the temporarily unavailable first storage for restoration purposes. The second type of backup may be chained to the previous backup stored in the temporarily unavailable storage. The second type of backup may be converted to the first type of backup when the first storage becomes available again in a computationally efficient manner.
1. A manager of a backup storage system, comprising:
persistent storage for storing metadata that associates clients with storages, wherein the metadata specifies that a backup for a client of the clients is stored on a first storage, wherein the backup storage system comprises the manager, the first storage, and a second storage; and
a data manager programmed to:
initiate, using the metadata, a backup generation for the client of the clients, wherein initiating the backup generation comprises sending a message to an agent on the client to initiate generation of an incremental backup;
obtain the incremental backup from the agent;
after the backup is initiated, make a determination that the first storage is unavailable; and
in response to the determination:
obtain an identity of a previous backup from the client, wherein the previous backup is one selected from a group consisting of a full backup and a synthetic backup;
generate a chain-based backup using the incremental backup and the identity of the previous backup; and
store the chain-based backup for the client in the second storage of the storages.
2. The manager of claim 1 , wherein no other backups for the client are stored to complete the backup generation.
3. The manager of claim 1 , wherein the data manager is further programmed to:
make a second determination, after storing the chain-based backup, that the first storage is available;
in response to the second determination:
generate a new synthetic backup for the client by merging the chain-based backup with a backup stored in the first storage.
4. The manager of claim 3 , wherein the backup is a full backup.
5. The manager of claim 4 , wherein the full backup is based on data of the client at a point in time, wherein the chain-based backup is based on changes to the data of the client over a period of time.
6. The manager of claim 3 , wherein the backup is an existing synthetic backup.
7. The manager of claim 6 , wherein the existing synthetic backup is based on an existing full backup and an existing chain-based backup.
8. A method for operating a manager of a backup storage system, comprising:
initiating, using metadata, a backup generation for a client of clients, wherein initiating the backup generation comprises sending a message to an agent on the client to initiate generation of an incremental backup, wherein the metadata specifies that a backup for the client of the clients is stored on a first storage, wherein the backup storage system comprises the manager, the first storage, and a second storage;
obtaining the incremental backup from the agent;
after initiating the backup, making a determination that the first storage is unavailable;
in response to the determination:
obtaining an identity of a previous backup from the client, wherein the previous backup is one selected from a group consisting of a full backup and a synthetic backup;
generating a chain-based backup using the incremental backup and the identity of the previous backup; and
storing the chain-based backup for the client in the second storage of the storages.
9. The method of claim 8 , wherein no other backups for the client are stored to complete the backup generation.
10. The method of claim 8 , further comprising:
making a second determination, after storing the chain-based backup, that the first storage is available;
in response to the second determination:
generating a new synthetic backup for the client by merging the chain-based backup with a backup stored in the first storage.
11. The method of claim 10 , wherein the backup is a full backup.
12. The method of claim 11 , wherein the full backup is based on data of the client at a point in time, wherein the chain-based backup is based on changes to the data of the client over a period of time.
13. The method of claim 10 , wherein the backup is an existing synthetic backup.
14. The method of claim 13 , wherein the existing synthetic backup is based on an existing full backup and an existing chain-based backup.
15. A non-transitory computer readable medium comprising computer readable program code, which when executed by a computer processor enables the computer processor to perform a method for operating a manager for a backup storage system, the method comprising:
initiating, using metadata, a backup generation for a client of clients, wherein initiating the backup generation comprises sending a message to an agent on the client to initiate generation of an incremental backup, wherein the metadata specifies that a backup for the client of the clients is stored on a first storage, wherein the backup storage system comprises the manager, the first storage, and a second storage;
obtaining the incremental backup from the agent;
after initiating the backup, making a determination that the first storage is unavailable;
in response to the determination:
obtaining an identity of a previous backup from the client, wherein the previous backup is one selected from a group consisting of a full backup and a synthetic backup;
generating a chain-based backup using the incremental backup and the identity of the previous backup;
storing the chain-based backup for the client in a second storage of the storages.
16. The non-transitory computer readable medium of claim 15 , wherein no other backups for the client are stored to complete the backup generation.
17. The non-transitory computer readable medium of claim 15 , wherein the method further comprises:
making a second determination, after storing the chain-based backup, that the first storage is available;
in response to the second determination:
generating a new synthetic backup for the client by merging the chain-based backup with a backup stored in the first storage.
18. The non-transitory computer readable medium of claim 17 , wherein the backup is a full backup.
19. The non-transitory computer readable medium of claim 18 , wherein the full backup is based on data of the client at a point in time, wherein the chain-based backup is based on changes to the data of the client over a period of time.
20. The non-transitory computer readable medium of claim 17 , wherein the backup is an existing synthetic backup.