Incremental reference counting validation based on a copy-on-write indexing structure

A method, comprising:

• maintaining, by a system comprising at least one processor, a mapping between tree page identifiers of a copy-on-write b+ tree and first in-use chunk identifiers of a storage system that are used to store data;
• performing, by the system, a reference count validation, comprising iterating over the mapping, to produce a validation result, wherein the validation result indicates whether respective chunks that correspond to the first in-use chunk identifiers are suitable for garbage collection because they are not pointed to by a data object, and wherein the iterating over the mapping comprises, for respective tree page identifiers of the tree page identifiers,
• based on a first tree page identifier of the tree page identifiers being present in the mapping,
• loading, by the system and from the mapping, second in-use chunk identifiers that correspond to the first tree page identifier, and
• updating, by the system, the validation result based on the loading of the second in-use chunk identifiers, and
• based on the first tree page identifier being omitted from the mapping,
• loading, by the system, content that corresponds to the first tree page identifier,
• updating, by the system, the validation result based on the extracting of the third in-use chunk identifiers, and
• updating, by the system, the mapping based on third in-use chunk identifiers from object metadata of the content that corresponds to the first tree page identifier, to produce an updated mapping;
• and
• performing, by the system, garbage collection of the storage system based on the validation result.

A non-transitory computer-readable medium comprising instructions that, in response to execution, cause a system comprising at least one processor to perform operations, comprising:

• performing a reference count validation, comprising iterating over a mapping between tree page identifiers of a copy-on-write b+ tree and first in-use chunk identifiers of a storage system that are used to store data, to produce a validation result, wherein the validation result indicates whether respective chunks that correspond to the first in-use chunk identifiers are suitable for garbage collection because they are not pointed to by a data object, and wherein the iterating over the mapping comprises, for respective tree page identifiers of the tree page identifiers,
• based on a first tree page identifier of the tree page identifiers being present in the mapping,
• loading, by the system and from the mapping, second in-use chunk identifiers that correspond to the first tree page identifier, and
• updating the validation result based on the loading of the second in-use chunk identifiers, and
• based on the first tree page identifier being omitted from the mapping,
• loading content that corresponds to the first tree page identifier,
• updating the validation result based on the extracting of the third in-use chunk identifiers, and
• updating the mapping based on third in-use chunk identifiers from object metadata of the content that corresponds to the first tree page identifier, to produce an updated mapping;
• and
• performing garbage collection of the storage system based on the validation result.