Abstract: Techniques are disclosed for ensuring consistent metadata across computing devices. In one example, a user device of a plurality of user devices receives a manifest that includes first metadata associated with a file system update of a file system of the user device. The user device generates second metadata of the file system based on performing the file system update. The user device then generates a dictionary based on comparing metadata records of the first metadata with metadata records of the second metadata. The dictionary may indicate a difference between at least one metadata record of the first metadata and at least one metadata record of the second metadata. The user device then updates the second metadata of the file system to match the first metadata based at least in part on the difference indicated by the dictionary.

Abstract: Techniques are disclosed for ensuring consistent metadata across computing devices. In one example, a user device of a plurality of user devices receives a manifest that includes first metadata associated with a file system update of a file system of the user device. The user device generates second metadata of the file system based on performing the file system update. The user device then generates a dictionary based on comparing metadata records of the first metadata with metadata records of the second metadata. The dictionary may indicate a difference between at least one metadata record of the first metadata and at least one metadata record of the second metadata. The user device then updates the second metadata of the file system to match the first metadata based at least in part on the difference indicated by the dictionary.

Abstract: Techniques are disclosed for ensuring consistent metadata across computing devices. In one example, a user device of a plurality of user devices receives a manifest that includes first metadata associated with a file system update of a file system of the user device. The user device generates second metadata of the file system based on performing the file system update. The user device then generates a dictionary based on comparing metadata records of the first metadata with metadata records of the second metadata. The dictionary may indicate a difference between at least one metadata record of the first metadata and at least one metadata record of the second metadata. The user device then updates the second metadata of the file system to match the first metadata based at least in part on the difference indicated by the dictionary.

Abstract: The embodiments set forth a technique for over-provisioning storage space within a solid-state storage device (SSD). In particular, a file system can (1) receive a first request to create a file, where the first request includes a size for the file, (2) identifying at least one extent that corresponds to storage space within the SSD that satisfies the size for the file, and associating the file with the at least one extent to indicate that the storage space is occupied, (3) receive a second request to cause (i) the file to remain established within the file system, and (ii) the storage space to be marked free within the SSD, and (4) carrying out the second request by causing the storage space to be marked free within the SSD.

Abstract: A device implementing a system for defragmenting metadata of a filesystem includes a processor configured to, in response to receiving a trigger from a server remote from the device, obtain the metadata from a first data structure, the first data structure comprising a first set of one or more nodes and a second set of one or more nodes, and insert the metadata obtained from the first data structure into a third set of one or more nodes of a second data structure, wherein the third set of one or more nodes omits one or more entries from the second set of nodes. The at least one processor is further configured to, in accordance with a determination that the metadata was successfully inserted into the second data structure, provide the second data structure as a replacement of the first data structure for the filesystem.

Abstract: The embodiments set forth a technique for over-provisioning storage space within a solid-state storage device (SSD). In particular, a file system can (1) receive a first request to create a file, where the first request includes a size for the file, (2) identifying at least one extent that corresponds to storage space within the SSD that satisfies the size for the file, and associating the file with the at least one extent to indicate that the storage space is occupied, (3) receive a second request to cause (i) the file to remain established within the file system, and (ii) the storage space to be marked free within the SSD, and (4) carrying out the second request by causing the storage space to be marked free within the SSD.