Snapshot management in distributed file systems

1. A method comprising:

presenting to a user, by a processing device, a list of snapshots of a distributed file system, wherein the snapshots comprise one or more active snapshots and one or more inactive snapshots, and wherein:

the one or more active snapshots are associated with one or more running processes comprising a data-serving process; and

the one or more inactive snapshots are not associated with one or more running processes;

receiving, by the processing device, a request by the user to access one of the snapshots in the list;

in response to determining that the requested snapshot is active, providing, to the user, access to the requested snapshot; and

in response to determining that the requested snapshot is inactive:

activating the requested snapshot by starting a set of processes comprising the data-serving process and associating the set of processes with the requested snapshot,

providing to the user, after the activating, access to the requested snapshot; and

in response to determining, after the activating, that a number of active snapshots exceeds a threshold number of active snapshots, selecting an active snapshot and deactivating the selected active snapshot, wherein the threshold number of active snapshots corresponds to a limit on resources of the distributed file system that are devoted to active snapshots.

2. The method of claim 1 wherein the presenting is in response to the user navigating to a designated subdirectory of a directory and submitting a command to list the contents of the designated subdirectory and wherein the designated subdirectory is a virtual subdirectory.

3. The method of claim 2 wherein each snapshot of the list of snapshots is presented as a respective subdirectory of the virtual subdirectory, and wherein providing access to a snapshot in the list of snapshots enables the user to navigate to the respective subdirectory.

4. The method of claim 1 wherein the data-serving process serves data associated with snapshots of the distributed file system and a second process that serves data associated with files of the distributed file system.

5. The method of claim 1 wherein the selecting is in view of a measure of how frequently one or more of the snapshots are accessed during a prior time interval.

6. The method of claim 1 wherein the user is associated with a client machine, and wherein accessing a file of the distributed file system comprises transmitting a first message from the client machine to a first server machine and a second message from the client machine to a second server machine, and wherein accessing a snapshot of the distributed file system comprises transmitting a third message from the client machine to the first server machine and a fourth message from the first server machine to the second server machine.

7. The method of claim 1 , wherein:

the one or more running processes further comprise a snapshot-management process, and

activating the requested snapshot further comprises starting the snapshot-management process and associating the snapshot-management process with the requested snapshot.

8. An apparatus comprising:

one or more storage devices to store data associated with a distributed file system; and

a processing device, operatively coupled to the one or more storage devices, to:

present to a user a list of snapshots of the distributed file system, wherein the snapshots comprise one or more active snapshots and one or more inactive snapshots, and wherein:

the one or more active snapshots are associated with one or more running processes comprising a data-serving process; and

the one or more inactive snapshots are not associated with one or more running processes;

receive a request by the user to access a first snapshot that is in the list;

in response to determining that the first snapshot is active, provide, to the user, access to the first snapshot; and

in response to determining that the first snapshot is inactive:

determine whether activation of the first snapshot will cause a number of active snapshots to exceed a threshold number of active snapshots, wherein the threshold number of active snapshots corresponds to a limit on resources of the distributed file system that are devoted to active snapshots;

in response to determining that the activation will cause the number of active snapshots to exceed the threshold number of active snapshots, select a second snapshot that is active and deactivate the selected second active snapshot;

activate the first snapshot by starting a set of processes comprising the data-serving process and associating the set of processes with the requested snapshot; and

provide to the user, after the activating of the first snapshot, access to the first snapshot.

9. The apparatus of claim 8 wherein the presenting is in response to the user navigating to a designated subdirectory of a directory and submitting a command to list the contents of the designated subdirectory and wherein the designated subdirectory is a virtual subdirectory.

10. The apparatus of claim 9 wherein each snapshot of the list of snapshots is presented as a respective subdirectory of the virtual subdirectory, and wherein providing access to a snapshot in the list of snapshots enables the user to navigate to the respective subdirectory.

11. The apparatus of claim 8 wherein the data-serving process serves data associated with snapshots of the distributed file system and a second process that serves data associated with files of the distributed file system.

12. The apparatus of claim 8 wherein the selecting is in view of a measure of how frequently one or more of the snapshots are accessed during a prior time interval.

13. The apparatus of claim 8 wherein the user is associated with a client machine, and wherein accessing a file of the distributed file system comprises transmitting a first message from the client machine to a first server machine and a second message from the client machine to a second server machine, and wherein accessing a snapshot of the distributed file system comprises transmitting a third message from the client machine to the first server machine and a fourth message from the first server machine to the second server machine.

14. A non-transitory computer readable storage medium, having instructions stored therein, which when executed, cause a processing device to:

present to a user, by the processing device, a list of snapshots of a distributed file system, wherein the snapshots comprise one or more active snapshots and one or more inactive snapshots and wherein:

the one or more active snapshots are associated with one or more running processes comprising a data-serving process; and

the one or more inactive snapshots are not associated with one or more running processes;

receive, by the processing device, a request by the user to access one of the snapshots in the list;

in response to determining that the requested snapshot is active, provide, to the user, access to the requested snapshot; and

in response to determining that the requested snapshot is inactive:

determine whether activation of the first snapshot will cause a number of active snapshots to exceed a threshold number of active snapshots, wherein the threshold number of active snapshots corresponds to a limit on resources of the distributed file system that are devoted to active snapshots;

in response to determining that the activation will cause the number of active snapshots to exceed the threshold number of active snapshots, select a second snapshot that is active and deactivate the selected second active snapshot and

activate the requested snapshot by starting a set of processes comprising the data-serving process and associating the set of processes with the requested snapshot and provide access to the requested snapshot to the user after the activating.

15. The non-transitory computer readable storage medium of claim 14 wherein the data-serving process serves data associated with snapshots of the distributed file system and a second process that serves data associated with files of the distributed file system.

16. The non-transitory computer readable storage medium of claim 14 wherein the presenting is in response to the user navigating to a designated subdirectory of a directory and submitting a command to list the contents of the designated subdirectory.

17. The non-transitory computer readable storage medium of claim 16 wherein the designated subdirectory is a virtual subdirectory, and wherein each snapshot of the list of snapshots is presented as a respective subdirectory of the virtual subdirectory, and wherein providing access to a snapshot in the list of snapshots enables the user to navigate to the respective subdirectory.

18. The non-transitory computer readable storage medium of claim 14 wherein the user is associated with a client machine, and wherein accessing a file of the distributed file system comprises transmitting a first message from the client machine to a first server machine and a second message from the client machine to a second server machine, and wherein accessing a snapshot of the distributed file system comprises transmitting a third message from the client machine to the first server machine and a fourth message from the first server machine to the second server machine.