Pushing a point in time to a backend object storage for a distributed storage system
A plurality of computing devices are communicatively coupled to each other via a network, and each of the plurality of computing devices is operably coupled to one or more of a plurality of storage devices. The computing devices may push a point in time to a backend for a distributed storage system.
1. A method for a distributed storage system, comprising:
taking a snapshot of a file system, distributed across multiple storage nodes, of a first data center, wherein the first data center is an active data center, and wherein:
the file system comprises a plurality of nodes and a plurality of bucket groups,
each node of the plurality of nodes is operable to host one or more buckets,
each bucket group of the plurality of bucket groups comprises a unique combination of nodes of the plurality of nodes, and
no two buckets hosted by the plurality of nodes share the same bucket group;
storing the snapshot to a bucket in an object storage; and
indicating to a second data center that the second data center is the active data center.
2. The method of claim 1 , wherein the file system is distributed across five nodes.
3. The method of claim 1 , wherein the first data center makes the file system read only prior to taking the snapshot.
4. The method of claim 3 , wherein the snapshot is a final snapshot by the first data center prior to going to a non-active data center status.
5. The method of claim 4 , wherein at a first pre-determined period before going to the non-active data center status, one or more snapshots are taken prior to the final snapshot.
6. The method of claim 5 , wherein the one or more snapshots are stored in the object storage.
7. The method of claim 1 , wherein the first data center stores its active system state to the object storage.
8. The method of claim 1 , wherein the distributed storage system is used for one of: cloud burst configuration or follow the sun configuration.
9. A method for a distributed storage system, comprising:
storing a snapshot of a file system, distributed across multiple storage nodes, to a bucket in a backend object storage, comprising:
traversing a snapshot configuration tree to determine changes to the file system;
reading an on-disk hash (ODH) table to enumerate changes to the ODH table;
putting the changes to the file system and the changes to the ODH table in one or more buffers; and
storing the buffer in the backend object storage, wherein:
the file system comprises a plurality of nodes and a plurality of bucket groups,
each node of the plurality of nodes is operable to host one or more buckets,
each bucket group of the plurality of bucket groups comprises a unique combination of nodes of the plurality of nodes, and
no two buckets hosted by the plurality of nodes share the same bucket group.
10. The method of claim 9 , wherein the file system is distributed across five nodes.
11. The method of claim 9 , comprising storing a final object to the backend object storage, wherein the final object comprises a number of the one or more buffers.
12. The method of claim 9 , wherein each of the one or more buffers comprises a name of the snapshot and an index of the each of the one or more buffers.
13. The method of claim 9 , wherein the snapshot configuration tree is traversed to find a first snapshot.
14. The method of claim 9 , wherein reading the ODH table comprises looking at later ODHs to determine whether there is a more up-to-date change for an object.
15. The method of claim 14 , wherein when the more up-to-date change for an object is found, only the more up-to-date change is stored in the backend object storage.
16. The method of claim 9 , wherein at least one data file is aggregated into at least one of the one or more buffers when the at least one data file is less than a pre-determined size.
17. A distributed storage system comprising:
a file system distributed across multiple storage nodes, wherein:
the file system comprises a plurality of nodes and a plurality of bucket groups,
each node of the plurality of nodes is operable to host one or more buckets,
each bucket group of the plurality of bucket groups comprises a unique combination of nodes of the plurality of nodes, and
no two buckets hosted by the plurality of nodes share the same bucket group; and
a processor, wherein the processor is configured to store a snapshot of the file system by:
traversing a snapshot configuration tree to determine changes to the file system;
reading an on-disk hash (ODH) table to enumerate changes to the ODH table;
putting the changes to the file system and the changes to the ODH table in one or more buffers; and
storing the buffer in a backend object storage that comprises a bucket.
18. The system of claim 17 , wherein the processor is configured to store a final object to the backend object storage, wherein the final object comprises a number of the one or more buffers.
19. The system of claim 17 , wherein the processor is configured to traverse the snapshot configuration tree to find a first snapshot.
20. The system of claim 17 , wherein the processor is configured to aggregate at least one data file into at least one of the one or more buffers when the at least one data file is less than a pre-determined size.