Abstract: Methods, computer program products, computer systems, and the like are disclosed that provide for the tracking of access patterns of inodes, and the issuing of inode read-ahead instructions, in order to pre-fetch inodes. Such a method can include, for example, identifying a unit of metadata in a file system, identifying a file system structure in the file system, determining whether a file structure of the file system structure is non-sequential, and, in response to a determination that the file structure is non-sequential, retrieving a list of units of metadata. In such embodiments, the file system structure is associated with the unit of metadata, and the determining includes accessing the file system structure. Further, in certain embodiments, the units of metadata identified in the list of units of metadata are stored in a storage device of the computer system.

Abstract: Disclosed herein are methods, systems, and processes to track access patterns of inodes, and to issue read-ahead instructions to pre-fetch inodes into memory. A location of a unit of metadata in a metadata storage area is determined. Another location in the metadata storage area that corresponds to a current metadata read operation is determined. Whether a metadata read-ahead operation can be performed is determined using the location of the unit of metadata and the another location. In response to a determination that the metadata read-ahead operation can be performed, the metadata-ahead operation is issued.

Abstract: Disclosed herein are methods, systems, and processes track access patterns of inodes, and to issue read-ahead instructions to pre-fetch inodes into memory. An inode is accessed and a directory of the inode is determined. Also determined is whether an entry for the directory exists in a global inode list. If the entry exists in the global inode list, whether a file structure of the directory is sequential or non-sequential is determined. If the entry does not exist in the global inode list, a new entry for the directory is added in the global inode list.

Abstract: Disclosed herein are methods, systems, and processes to track access patterns of inodes, and to issue read-ahead instructions to pre-fetch inodes into memory. A location of a unit of metadata in a metadata storage area is determined. Another location in the metadata storage area that corresponds to a current metadata read operation is determined. Whether a metadata read-ahead operation can be performed is determined using the location of the unit of metadata and the another location. In response to a determination that the metadata read-ahead operation can be performed, the metadata-ahead operation is issued.

Abstract: Disclosed herein are methods, systems, and processes track access patterns of inodes, and to issue read-ahead instructions to pre-fetch inodes into memory. An inode is accessed and a directory of the inode is determined. Also determined is whether an entry for the directory exists in a global inode list. If the entry exists in the global inode list, whether a file structure of the directory is sequential or non-sequential is determined. If the entry does not exist in the global inode list, a new entry for the directory is added in the global inode list.

Abstract: A method of reading data from a partitioned directory incident to a serialized process. A first read and an offset value are received. A first data block in a modeled fully partitioned directory is identified based on the offset value and a predetermined number of entries associated with a buffer. It is determined whether the first data block in the fully partitioned directory is present in the actual partitioned directory. Zeros are written in the buffer if the first data block in the fully partitioned directory is not present in the actual partitioned directory otherwise the first data block associated with the actual partitioned directory is written to the buffer. A second data block is similarly read by a second read operation and written. The second data block is associated with a second subdirectory, selected based on a horizontal node traversal at a node level of said first subdirectory.

Abstract: A method of reading data from a partitioned directory incident to a serialized process. A first read and an offset value are received. A first data block in a modeled fully partitioned directory is identified based on the offset value and a predetermined number of entries associated with a buffer. It is determined whether the first data block in the fully partitioned directory is present in the actual partitioned directory. Zeros are written in the buffer if the first data block in the fully partitioned directory is not present in the actual partitioned directory otherwise the first data block associated with the actual partitioned directory is written to the buffer. A second data block is similarly read by a second read operation and written. The second data block is associated with a second subdirectory, selected based on a horizontal node traversal at a node level of said first subdirectory.

Abstract: A method of reading data from a partitioned directory incident to a serialized process. A first read and an offset value are received. A first data block in a modeled fully partitioned directory is identified based on the offset value and a predetermined number of entries associated with a buffer. It is determined whether the first data block in the fully partitioned directory is present in the actual partitioned directory. Zeros are written in the buffer if the first data block in the fully partitioned directory is not present in the actual partitioned directory otherwise the first data block associated with the actual partitioned directory is written to the buffer. A second data block is similarly read by a second read operation and written. The second data block is associated with a second subdirectory, selected based on a horizontal node traversal at a node level of said first subdirectory.

Abstract: A method of reading data from a partitioned directory incident to a serialized process. A first read and an offset value are received. A first data block in a modeled fully partitioned directory is identified based on the offset value and a predetermined number of entries associated with a buffer. It is determined whether the first data block in the fully partitioned directory is present in the actual partitioned directory. Zeros are written in the buffer if the first data block in the fully partitioned directory is not present in the actual partitioned directory otherwise the first data block associated with the actual partitioned directory is written to the buffer. A second data block is similarly read by a second read operation and written. The second data block is associated with a second subdirectory, selected based on a horizontal node traversal at a node level of said first subdirectory.