Data access during data recovery
A method, article of manufacture, and apparatus for accessing data during data recovery. In some embodiments, this includes sending an I/O request from an application to an object, wherein the object is being recovered, establishing an I/O intercept, intercepting the application's I/O request with the I/O intercept, and redirecting the I/O request based on the status of the object's sub-objects.
1. A method, comprising:
intercepting a request to access an object that is being recovered as part of a data recovery operation currently being performed, wherein the object includes one or more sub-objects;
in response to intercepting the access request, determining a recovery status for a first sub-object of the one or more sub-objects in relation to the data recovery operation;
determining, based at least in part on the determined recovery status of the first sub-object, an access method for accessing at least one of the one or more sub-objects in connection with the request to access the object, wherein determining whether the first sub-object is to be accessed via local I/O, remote I/O, or recovery on access based at least in part on the recovery status for the first sub-object, and wherein the determining of the access method is performed in connection with the recovery operation currently being performed;
in response to determining that the first sub-object is available via local I/O and that I/O has been successfully performed on a local copy of the first sub-object, accessing the local copy of the first sub-object; and
in response to determining that the first sub-object is available via local I/O and that I/O has not been performed on the local copy of the first sub-object, rerouting, an I/O request to access the object to a recovery service that manages recovery of the object.
2. The method of claim 1 , wherein the determining the access method further includes determining whether to change a priority of the first sub-object.
3. The method of claim 1 , wherein the request to access the object is intercepted by an I/O intercept, and wherein the I/O intercept is configured to apply the determined access method and transparently redirect the I/O request.
4. The method of claim 3 , wherein in response to determining that the first sub-object is available via local I/O and that I/O has been successfully performed on the local copy of the first sub-object, the I/O intercept is abolished with regard to the local copy of the first sub-object after the I/O has been successfully performed on the local copy of the first sub-object.
5. The method of claim 3 , wherein the I/O request to access the object is rerouted by the I/O intercept.
6. The method of claim 1 , wherein the determining the recovery status of the first sub-object includes determining whether the first sub-object is available via local I/O, remote I/O, or recovery on access.
7. The method of claim 6 , wherein in response to determining that the first sub-object is available via local I/O, the access method for an application to access the object includes redirecting the I/O request to a local copy of the first sub-object.
8. The method of claim 6 , wherein in response to determining that the first sub-object is available via remote I/O, the access method for an application to access the object includes redirecting the I/O request to a remote copy of the first sub-object.
9. The method of claim 8 , wherein the first sub-object is available via remote access and the I/O request is redirected to a remote copy of the first sub-object based at least in part on a pinned status of at least the first sub-object, and wherein pinning at least the first sub-object restricts access to the first sub-object to remote I/O.
10. The method of claim 6 , wherein in response to determining that the first sub-object is available via recovery on access, the access method for an application to access the object includes holding the I/O request until the first sub-object has been recovered.
11. The method of claim 6 , further comprising spinning up at least the first sub-object on a system and establishing a connection between an application and the spun-up first sub-object.
12. The method of claim 11 , wherein the system comprises a virtual machine mounted on top of a physical machine that is connected by a comparatively high bandwidth connection to a storage container storing the first sub-object.
13. The method of claim 1 , wherein the object is segmented into the one or more sub-objects based at least in part on one or more of the following: a predefined segment interval, an access pattern associated with the object, metadata, one or more policies, and one or more user commands.
14. The method of claim 1 , wherein each of at least a subset of the one or more sub-objects is assigned a recovery priority based at least in part on one or more of the following: an access pattern associated with the object, metadata, one or more polices, and one or more user commands.
15. The method of claim 14 , wherein the recovery priority corresponding to at least one of the one or more sub-objects is dynamically changed during the data recovery operation.
16. The method of claim 14 , wherein the recovery priority corresponding to the at least one of the one or more sub-objects is dynamically changed during the data recovery operation based on one or more access requests for the at least one of the one or more sub-objects.
17. The method of claim 1 , wherein each of at least a subset of the one or more sub-objects is assigned a recovery priority, wherein the recovery priority is stored in metadata or a header associated with the subset of the one or more sub-objects.
18. The method of claim 17 , wherein the recovery priority for at least one of the one or more sub-objects is updated during the recovery operation based at least in part on an access request for accessing the at least one of the one or more sub-objects.
19. A system for accessing data during data recovery, comprising a storage device and a processor configured to:
intercept a request to access an object that is being recovered as part of a data recovery operation currently being performed, wherein the object includes one or more sub-objects;
in response to intercepting the access request, determine a recovery status for a first sub-object of the one or more sub-objects in relation to the data recovery operation;
determine, based at least in part on the determined recovery status of the first sub-object, an access method for accessing at least one of the one or more sub-objects in connection with the request to access the object, wherein determining whether the first sub-object is to be accessed via local I/O, remote I/O, or recovery on access based at least in part on the recovery status for the first sub-object, and wherein the determining of the access method is performed in connection with the recovery operation currently being performed;
in response to determining that the first sub-object is available via local I/O and that I/O has been successfully performed on a local copy of the first sub-object, access the local copy of the first sub-object; and
in response to determining that the first sub-object is available via local I/O and that I/O has not been performed on the local copy of the first sub-object, reroute, an I/O request to access the object to a recovery service that manages recovery of the object.
20. A computer program product for accessing data during data recovery, comprising a non-transitory computer usable medium having machine readable code embodied therein for:
intercepting a request to access an object that is being recovered as part of a data recovery operation currently being performed, wherein the object includes one or more sub-objects;
in response to intercepting the access request, determining a recovery status for a first sub-object of the one or more sub-objects in relation to the data recovery operation;
determining, based at least in part on the determined recovery status of the first sub-object, an access method for accessing at least one of the one or more sub-objects in connection with the request to access the object, wherein determining whether the first sub-object is to be accessed via local I/O, remote I/O, or recovery on access based at least in part on the recovery status for the first sub-object, and wherein the determining of the access method is performed in connection with the recovery operation currently being performed;
in response to determining that the first sub-object is available via local I/O and that I/O has been successfully performed on a local copy of the first sub-object, accessing the local copy of the first sub-object; and
in response to determining that the first sub-object is available via local I/O and that I/O has not been performed on the local copy of the first sub-object, rerouting, an I/O request to access the object to a recovery service that manages recovery of the object.