Namespace salvage in logging-based filesystems

A system comprising:

• a processor;
• and memory configured to store one or more sequences of instructions which, when executed by the processor, cause the processor to carry out the steps of:
• scanning a linear representation of pages belonging to a tree structure that has been corrupted, the tree structure storing metadata of files managed by a logging-based filesystem, and comprising leaf pages having keys indexing the metadata;
• during the scanning, salvaging keys from leaf pages that are not corrupted by copying out the keys from uncorrupted leaf pages into a new tree structure;
• retrieving a plurality of snapshots that have been taken of the tree structure;
• repeating the scanning using the plurality of snapshots that have been taken of the tree structure to find keys that are missing from the new tree structure;
• upon finding a missing key in an uncorrupted leaf page of a snapshot taken of the tree structure, copying out the missing key into the new tree structure unless the missing key has already been found and copied out from an uncorrupted leaf page of a more recent snapshot taken of the tree structure;
• wherein the keys comprise a plurality of first keys and a plurality of second keys, each first key indexes an inode of a file, and each second key indexes a name of the file, and the processor further carries out the steps of:
• when a first key indexing an inode of a particular file has been found, but a second key indexing a name of the particular file has not been found, generating a new name for the particular file that includes an inode number of the particular file, and a placeholder name;
• assigning the new name to the particular file;
• reviewing the new tree structure to identify a directory within which the particular file resided at a previous point-in-time;
• and
• placing the particular file having the new name in the identified directory.

A computer program product, comprising a non-transitory computer-readable medium having a computer-readable program code embodied therein, the computer-readable program code adapted to be executed by one or more processors to implement a method comprising:

• scanning a linear representation of pages belonging to a tree structure that has been corrupted, the tree structure storing metadata of files managed by a logging-based filesystem, and comprising leaf pages having keys indexing the metadata;
• during the scanning, salvaging keys from leaf pages that are not corrupted by copying out the keys from uncorrupted leaf pages into a new tree structure;
• retrieving a plurality of snapshots that have been taken of the tree structure;
• repeating the scanning using the plurality of snapshots that have been taken of the tree structure to find keys that are missing from the new tree structure;
• upon finding a missing key in an uncorrupted leaf page of a snapshot taken of the tree structure, copying out the missing key into the new tree structure unless the missing key has already been found and copied out from an uncorrupted leaf page of a more recent snapshot taken of the tree structure;
• wherein the keys comprise a plurality of first keys and a plurality of second keys, each first key indexes an inode of a file, and each second key indexes a name of the file, and the method further comprises:
• when a first key indexing an inode of a particular file has been found, but a second key indexing a name of the particular file has not been found, generating a new name for the particular file that includes an inode number of the particular file, and a placeholder name;
• assigning the new name to the particular file;
• reviewing the new tree structure to identify a directory within which the particular file resided at a previous point-in-time;
• and
• placing the particular file having the new name in the identified directory.