Sustaining backup service level objectives using dynamic resource allocation
A backup engine of a first storage system receives a request to perform a backup session from the first storage system to a second storage system based on a backup service level objective (SLO) that has been configured for the backup session. In response to the request, it is determined that a first backup resource allocated for the backup session by the first storage system cannot satisfy the SLO based on statistics of prior backup sessions in view of characteristics of the backup session to be performed. A dynamic resource allocation (DRA) module is to dynamically perform a first DRA to modify the first backup resource to satisfy the SLO. The backup engine then initiates the backup session by transmitting backup data from the first storage system to the second storage system using the modified first backup resource.
1. A computer-implemented method for performing a backup session, the method comprising:
receiving, by a backup engine of a first storage system, a request to perform a backup session from the first storage system to a second storage system based on a backup service level objective (SLO) that has been configured for the backup session, wherein the backup SLO includes information specifying that the backup session has to be performed within a predetermined period of time;
in response to the request, determining, prior to performing the backup session, that a first backup resource allocated for the backup session by the first storage system cannot satisfy the SLO based on statistics of prior backup sessions in view of characteristics of the backup session to be performed, wherein determining that a first backup resource allocated for the backup session by the first storage system cannot satisfy the SLO comprises: determining a backup size of the backup session by calculating a number of data pages that have been changed since a last backup session from the first storage system to the second storage system; and determining that the backup session cannot be completed within the predetermined period of time based on the backup size in view of the statistics of the prior backup sessions;
dynamically performing, by a dynamic resource allocation (DRA) module hosted by the first storage system, a first DRA to modify the first backup resource to satisfy the SLO; and
initiating the backup session by transmitting backup data from the first storage system to the second storage system using the modified first backup resource.
2. The method of claim 1 , further comprising transmitting a notification from the first storage system to the second storage system to cause the second storage system to modify a second backup resource allocated for the backup session by the second storage system.
3. The method of claim 1 , further comprising:
monitoring performance of the backup session in view of the SLO;
determining during the backup session that the backup session still does not satisfy the SLO using the modified first backup resource; and
dynamically performing a second DRA to further modify the modified first backup resource.
4. The method of claim 3 , further comprising transmitting a notification from the first storage system to the second storage system to allow the second storage system to further modify the second backup resource, in response to determining that the backup session still does not satisfy the SLO.
5. The method of claim 1 , wherein performing a first DRA comprises allocating more processor and memory resources to be exclusively utilized by the backup session.
6. The method of claim 5 , further comprising:
allocating one or more additional input and output (IO) paths between the first storage system and the second storage system; and
deallocating one or more existing IO paths from a host device that has been allocated to the host device for accessing data stored in the first storage system.
7. The method of claim 5 , further comprising, prior to initiating the backup session, moving at least a portion of the backup data from a first storage device to a second storage device within the first storage system, wherein the second storage device has a higher speed than the first storage device.
8. The method of claim 1 , wherein performing a first DRA comprises modifying the SLO to be in alignment with the modified first backup resource for subsequent backup sessions.
9. A non-transitory machine-readable medium having instructions stored therein, which when executed by a processor, cause the processor perform operations performing a backup session, the operations comprising:
receiving, by a backup engine of a first storage system, a request to perform a backup session from the first storage system to a second storage system based on a backup service level objective (SLO) that has been configured for the backup session, wherein the backup SLO includes information specifying that the backup session has to be performed within a predetermined period of time;
in response to the request, determining, prior to performing the backup session, that a first backup resource allocated for the backup session by the first storage system cannot satisfy the SLO based on statistics of prior backup sessions in view of characteristics of the backup session to be performed, wherein determining that a first backup resource allocated for the backup session by the first storage system cannot satisfy the SLO comprises: determining a backup size of the backup session by calculating a number of data pages that have been changed since a last backup session from the first storage system to the second storage system; and determining that the backup session cannot be completed within the predetermined period of time based on the backup size in view of the statistics of the prior backup sessions;
dynamically performing, by a dynamic resource allocation (DRA) module hosted by the first storage system, a first DRA to modify the first backup resource to satisfy the SLO; and
initiating the backup session by transmitting backup data from the first storage system to the second storage system using the modified first backup resource.
10. The non-transitory machine-readable medium of claim 9 , wherein the operations further comprise transmitting a notification from the first storage system to the second storage system to cause the second storage system to modify a second backup resource allocated for the backup session by the second storage system.
11. The non-transitory machine-readable medium of claim 9 , wherein the operations further comprise:
monitoring performance of the backup session in view of the SLO;
determining during the backup session that the backup session still does not satisfy the SLO using the modified first backup resource; and
dynamically performing a second DRA to further modify the modified first backup resource.
12. The non-transitory machine-readable medium of claim 11 , wherein the operations further comprise transmitting a notification from the first storage system to the second storage system to allow the second storage system to further modify the second backup resource, in response to determining that the backup session still does not satisfy the SLO.
13. The non-transitory machine-readable medium of claim 9 , wherein performing a first DRA comprises allocating more processor and memory resources to be exclusively utilized by the backup session.
14. The non-transitory machine-readable medium of claim 13 , wherein the operations further comprise:
allocating one or more additional input and output (IO) paths between the first storage system and the second storage system; and
deallocating one or more existing IO paths from a host device that has been allocated to the host device for accessing data stored in the first storage system.
15. The non-transitory machine-readable medium of claim 13 , wherein the operations further comprise, prior to initiating the backup session, moving at least a portion of the backup data from a first storage device to a second storage device within the first storage system, wherein the second storage device has a higher speed than the first storage device.
16. The non-transitory machine-readable medium of claim 9 , wherein performing a first DRA comprises modifying the SLO to be in alignment with the modified first backup resource for subsequent backup sessions.
17. A data processing system operating as a first storage system, comprising:
a processor;
a memory coupled to the processor;
a backup engine coupled to the memory and the processor to
receive a request to perform a backup session from the first storage system to a second storage system based on a backup service level objective (SLO) that has been configured for the backup session, wherein the backup SLO includes information specifying that the backup session has to be performed within a predetermined period of time,
in response to the request, determine, prior to performing the backup session, that a first backup resource allocated for the backup session by the first storage system cannot satisfy the SLO based on statistics of prior backup sessions in view of characteristics of the backup session to be performed, wherein determining that a first backup resource allocated for the backup session by the first storage system cannot satisfy the SLO comprises: determining a backup size of the backup session by calculating a number of data pages that have been changed since a last backup session from the first storage system to the second storage system; and determining that the backup session cannot be completed within the predetermined period of time based on the backup size in view of the statistics of the prior backup sessions; and
a dynamic resource allocation (DRA) module to dynamically perform a first DRA to modify the first backup resource to satisfy the SLO, wherein the backup engine is to initiate the backup session by transmitting backup data from the first storage system to the second storage system using the modified first backup resource.
18. The system of claim 17 , wherein the DRA module transmits a notification from the first storage system to the second storage system to cause the second storage system to modify a second backup resource allocated for the backup session by the second storage system.
19. The system of claim 17 , further comprising a backup monitor configured to:
monitor performance of the backup session in view of the SLO; and
determine during the backup session that the backup session still does not satisfy the SLO using the modified first backup resource, wherein the DRA module is to dynamically perform a second DRA to further modify the modified first backup resource.