Tracking version families in a file system

1. A method of facilitating file system operations in a data storage system, the method comprising:

storing a VFDB (version family database) for each of multiple version families in a file system of the data storage system, each VFDB including information pertaining to a respective version family, each version family including a respective set of files in the file system that are related to one another by file snapping; and

upon creation of a snap of a file that belongs to one of the version families, storing an identifier of the snap together with an identifier of the file in a VFDB of the version family to which the file belongs,

wherein storing the VFDB for each version family includes storing a list of identifiers of all files that belong to that version family, wherein storing the VFDB for each of the version families includes (i) storing a first record in the VFDB, the first record including information pertaining to the respective version family as a whole and (ii) storing an additional record in the VFDB for each file that belongs to that version family, and

wherein the method further comprises:

receiving a command to delete a file in the file system;

reading a version family identifier from an inode of the file to be deleted;

looking up a VFDB using the version family identifier;

identifying another file listed in the VFDB; and

transferring reference counts of a set of block pointers of the file to be deleted to reference counts of block pointers of the other file identified in the VFDB.

2. The method of claim 1 , wherein the file system provides an inode for each file in the file system that belongs to a version family, and wherein the method further comprises storing a version family identifier in the inode of each such file, the version family identifier identifying the version family to which the respective file belongs.

3. The method of claim 2 , wherein the method further comprises:

accessing an inode of a file in the file system;

retrieving a version family identifier from the inode of that file; and

locating the VFDB of the version family to which that file belongs based on the version family identifier.

4. The method of claim 3 , wherein each VFDB is stored in a respective file in the file system, and wherein locating the VFDB includes:

generating a file name based on the version family identifier; and

performing a directory lookup for the file name in the file system.

5. The method of claim 1 , wherein storing the VFDB for a version family includes storing a set of counts of blocks allocated to the set of files in that version family, and wherein the method further comprises:

accessing the VFDB to obtain the set of counts; and

generating a family liability value for the version family, the family liability value based at least in part on the set of counts accessed from the VFDB.

6. The method of claim 5 , wherein storing the first record in the VFDB includes storing a count of blocks allocated to all files in the version family, wherein generating the family liability value for the version family is based at least in part on the count of blocks allocated to all files in the version family.

7. The method of claim 1 , further comprising:

receiving a command to move contents of a block from a first location in the file system to a second location in the file system;

identifying an inode of a file in the file system that points to the block;

reading a version family identifier from the inode of the file that points to the block;

looking up a VFDB using the version family identifier; and

for each file identified in the VFDB, (i) testing whether that file points to the block at the first location and, (ii) if it does, redirecting that file to point to the block at the second location.

8. The method of claim 1 , wherein the data storage system includes a first storage processor and a second storage processor, wherein a VFDB for a version family is accessible by both the first storage processor and the second storage processor, and wherein the method further comprises:

providing host access to files of a version family by a set of software objects running on the first storage processor;

upon a failure of the first storage processor, populating the set of software objects on the second storage processor by (i) accessing information about each file in the version family from the VFDB and (ii) providing the accessed information to the set of software objects on the second storage processor; and

providing host access to the files of the version family by the set of software objects running on the second storage processor.

9. The method of claim 1 , further comprising, in response to the data storage system receiving an IO (input/output) request that specifies an allocating write to a file in a version family:

updating at least one counter of allocated blocks in an inode of the file; and

updating at least one counter of allocated blocks in a VFDB of the version family.

10. The method of claim 9 , wherein updating at least one counter of allocated blocks in the VFDB includes:

updating a count of blocks in the first record of the VFDB, the count of blocks in the first record of the VFSB counting blocks allocated to all files in the version family; and

updating a count of blocks in an additional record of the VFDB, the count of blocks in the additional record of the VFSB counting blocks allocated to the file.

11. The method of claim 1 , wherein storing the VFDB for a version family includes storing a set of counts of blocks allocated to the set of files in that version family, and wherein the method further comprises:

accessing the VFDB to obtain the set of counts;

generating a family liability value for the version family, the family liability value based at least in part on the set of counts accessed from the VFDB; and

obtaining insurance to cover the family liability value for the version family from a storage pool.

12. A computer program product including a set of non-transitory, computer-readable media having instructions which, when executed by control circuitry of a data storage system, cause the control circuitry to perform a method for managing storage space in a data storage system, the method comprising:

storing a VFDB (version family database) for each of multiple version families in a file system of the data storage system, each VFDB including information pertaining to a respective version family, each version family including a respective set of files in the file system that are related to one another by file snapping; and

upon creation of a snap of a file that belongs to one of the version families, storing an identifier of the snap together with an identifier of the file in a VFDB of the version family to which the file belongs,

wherein storing the VFDB for each version family includes (i) storing a list of identifiers of all files that belong to that version family, (ii) storing a first record in the VFDB, the first record including information pertaining to the respective version family as a whole and (iii) storing an additional record in the VFDB for each file that belongs to that version family,

wherein the file system provides an inode for each file in the file system that belongs to a version family, wherein the method further comprises storing a version family identifier in the inode of each such file, the version family identifier identifying the version family to which the respective file belongs; and

wherein storing the VFDB for a version family includes storing a set of counts of blocks allocated to the set of files in that version family, and wherein the method further comprises:

accessing the VFDB to obtain the set of counts; and

generating a family liability value for the version family, the family liability value based at least in part on the set of counts accessed from the VFDB.

13. The computer program product of claim 12 , wherein each VFDB is stored in a respective file in the file system, and wherein the method further comprises:

accessing an inode of a file in the file system;

retrieving a version family identifier from the inode of that file; and

locating the VFDB of the version family to which that file belongs based on the version family identifier,

wherein locating the VFDB includes generating a file name based on the version family identifier and performing a directory lookup for the file name in the file system.

14. The computer program product of claim 12 , wherein the method further comprises:

receiving a command to move contents of a block from a first location in the file system to a second location in the file system;

identifying an inode of a file in the file system that points to the block;

reading a version family identifier from the inode of the file that points to the block;

looking up a VFDB using the version family identifier; and

for each file identified in the VFDB, (i) testing whether that file points to the block at the first location and, (ii) if it does, redirecting that file to point to the block at the second location.

15. A computer program product including a set of non-transitory, computer-readable media having instructions which, when executed by control circuitry of a data storage system, cause the control circuitry to perform a method for managing storage space in a data storage system, the method comprising:

storing a VFDB (version family database) for each of multiple version families in a file system of the data storage system, each VFDB including information pertaining to a respective version family, each version family including a respective set of files in the file system that are related to one another by file snapping; and

upon creation of a snap of a file that belongs to one of the version families, storing an identifier of the snap together with an identifier of the file in a VFDB of the version family to which the file belongs,

wherein storing the VFDB for each version family includes (i) storing a list of identifiers of all files that belong to that version family, (ii) storing a first record in the VFDB, the first record including information pertaining to the respective version family as a whole and (iii) storing an additional record in the VFDB for each file that belongs to that version family,

wherein the file system provides an inode for each file in the file system that belongs to a version family, wherein the method further comprises storing a version family identifier in the inode of each such file, the version family identifier identifying the version family to which the respective file belongs; and

wherein the method further comprises, in response to the data storage system receiving an IO (input/output) request that specifies an allocating write to a file in a version family:

updating at least one counter of allocated blocks in an inode of the file; and

updating at least one counter of allocated blocks in a VFDB of the version family.