Fast migration of metadata

1. A system, comprising:

a processor configured to:

select one or more buckets of key-value pairs of a first node of a distributed storage system to migrate to a second node of the distributed storage system;

identify one or more underlying database files corresponding to the one or more selected buckets;

directly copy the one or more identified underlying database files from a storage of a first node to a storage of the second node; and

link the copied underlying database files in a database of the second node to implement the one or more selected buckets in the second node; and

a memory coupled to the processor and configured to provide the processor with instructions.

2. The system of claim 1 , wherein the one or more buckets of key-value pairs are stored is across a plurality of nodes of the distributed storage system.

3. The system of claim 1 , wherein the processor is further configured to log one or more updates to the one or more selected buckets while the one or more identified underlying database files are being directly copied to the second node.

4. The system of claim 3 , wherein to log the one or more updates to the one or more selected buckets while the one or more identified underlying database files are being directly copied to the second node, the processor is further configured to:

generate a first log to store one or more updates to the one or more selected buckets corresponding to the one or more underlying database files; and

inspect the first log after the one or more identified underlying database files have been directly copied to the second node.

5. The system of claim 4 , wherein to inspect the first log after the one or more identified underlying database files have been directly copied to the second node, the processor is further configured to compare a number of entries included in the first log to a threshold.

6. The system of claim 5 , wherein in response to determining that the number of entries included in the first log is less than the threshold, the processor is further configured to:

suspend updates to the one or more selected buckets; and

transfer to the second node the one or more updates stored in the first log.

7. The system of claim 5 , wherein in response to determining that the number of entries included in the first log is not less than the threshold, the processor is further configured to determine whether a retry threshold has been exceeded.

8. The system of claim 7 , wherein in response to determining that the retry threshold has been exceeded, the processor is further configured to:

suspend updates to the one or more selected buckets; and

transfer to the second node the one or more updates stored in a current log.

9. The system of claim 7 , wherein in response to determining that the retry threshold has been exceeded, the processor is further configured to:

generate a new log; and

log one or more updates associated with the one or more selected buckets in the new log.

10. The system of claim 9 , wherein the processor is further configured to transfer to the second node the new log.

11. The system of claim 1 , wherein the processor is further configured to update metadata to transfer a node assignment associated with the one or more selected buckets from the first node to the second node.

12. The system of claim 1 , wherein the second node is configured to:

receive the one or more underlying database files corresponding to the one or more selected buckets; and

update a database of the second node based on the one or more underlying database files.

13. The system of claim 12 , wherein the second node is further configured to:

receive one or more logs that include one or more updates to the one or more selected buckets; and

update the database of the second node based on the one or more received logs.

14. The system of claim 1 , wherein data associated with the one or more selected buckets is organized using a snapshot tree.

15. A method, comprising:

selecting one or more buckets of key-value pairs of a first node of a distributed storage system to migrate to a second node of the distributed storage system;

identifying one or more underlying database files corresponding to the one or more selected buckets;

directly copying the one or more identified underlying database files from a storage of the first node to a storage of the second node; and

linking the copied underlying database files in a database of the second node to implement the one or more selected buckets in the second node.

16. The method of claim 15 , wherein the one or more buckets of key-value pairs are stored across a plurality of nodes of the distributed storage system.

17. The method of claim 15 , further comprising logging one or more updates to the one or is more selected buckets while the one or more identified underlying database files are being directly copied to the second node.

18. The method of claim 17 , wherein logging the one or more to the one or more selected buckets while the one or more identified underlying database files are being directly copied to the second node includes:

generating a first log to store one or more updates to the one or more selected buckets corresponding to the one or more underlying database files; and

inspecting the first log after the one or more identified underlying database files have been directly copied to the second node.

19. The method of claim 18 , wherein inspecting the first log after the one or more identified underlying database files have been directly copied to the second node includes:

comparing a number of entries included in the first log to a threshold;

in response to determining that the number of entries included in the first log is less than the threshold, suspending updates to the one or more selected buckets; and

transferring to the second node the one or more updates stored in the first log.

20. A computer program product, the computer program product being embodied in a non-transitory computer readable storage medium and comprising computer instructions for:

selecting one or more buckets of key-value pairs of a first node of a distributed storage system to migrate to a second node of the distributed storage system;

identifying one or more underlying database files corresponding to the one or more selected buckets;

directly copying the one or more identified underlying database files from a storage of a first node to a storage of the second node; and

linking the copied underlying database files in a database of the second node to implement the one or more selected buckets in the second node.