摘要:
A network storage server includes a tool for detecting and fixing errors while the network storage server remains online (available for servicing client requests), which includes enabling a user to approve or disapprove remedial changes before the changes are committed. The technique bypasses the usual consistency point process for new or modified data blocks representing potential remedial changes. At a consistency point, dirty data blocks representing the potential remedial changes are written to a change log file residing outside the volume. The modified data blocks are written in sequential order to logical blocks of the change log file. In response to a user input indicating that a potential change should be committed, the corresponding modified data blocks are read from the change log file in the order in which they were written to the change log file, and they are written to persistent storage in that order.
摘要:
This disclosure presents checking file system consistency for a storage server. During a start up phase of a file system, a consistency checker, upon receiving a request from an internal client for a part of data managed by a file system of the storage server, checks file system consistency only for a part of metadata. The part of metadata is used to maintain consistency of the requested data. After the consistency check of the part of metadata is completed, the internal client is allowed to access the requested data before a remainder of the metadata is checked for consistency.
摘要:
The present invention relates to a system for restoring a file from a snapshot, where a version of the file exists in both an active file system and the snapshot. A twin inode is created in the active file system and comparisons are made between block pointers of the twin inode and the snapshot. If there is a match, the block pointer of the twin inode is moved to the active file system. If there is not a match, a determination is made whether the snapshot block pointer exists in the active file system. If the snapshot block pointer does not exist in the active file system, it is copied to the active file system. If it does exist, then the actual data block pointed to by the snapshot block pointer is copied to the active file system. In this way, a file may be restored without the need to always copy every individual data block or inode from the snapshot.
摘要:
A method for reducing duplicate data of an active file system, where the data is stored on a storage device of a storage system, is disclosed. A persistent image of the active file system is generated at a first point in time. At a later time, a sequence of steps to compare the active file system with the persistent image is initiated. Then, if a block of data of the active file system is identical to the data in the persistent image, the active file system is directed to refer to the persistent image for data of the file. Alternatively, if the block of data of the active file system is not identical to the data in the persistent image, the active file system is directed to refer to the data of the active file system. A file folding command may be issued to start the comparison.
摘要:
A system and method maintains mappings from data containers to their parent directories within a file system of a storage system. Each inode is modified to include a primary name data structure comprising a parent directory inode and a parent directory cookie value. The parent directory cookie value identifies a particular directory entry within a specified block of a directory identified by parent directory inode. An alternate name file is utilized to store alternate names, such as those associated with hard links.
摘要:
A system and method verifies and restores the consistency of inode to pathname mappings. In a first embodiment, an off-line verification procedure is modified to verify and correct the primary name inode to pathname mapping information within inodes of a file system. In a second embodiment, an on-line file system verification process is modified to verify inode to pathname mapping information upon the loading of each inode within the file system.
摘要:
A method and apparatus for a reliable data storage system using block level checksums appended to data blocks. Files are stored on hard disks in storage blocks, including data blocks and block-appended checksums. The block-appended checksum includes a checksum of the data block, a VBN, a DBN, and an embedded checksum for checking the integrity of the block-appended checksum itself. A file system includes file blocks with associated block-appended checksum to the data blocks. The file blocks with block-appended checksums are written to storage blocks. In a preferred embodiment a collection of disk drives are formatted with 520 bytes of data per sector. For each 4,096-byte file block, a corresponding 64-byte block-appended checksum is appended to the file block with the first 7 sectors including most of the file block data while the 8th sector includes the remaining file block data and the 64-byte block-appended checksum.
摘要:
The invention provides a method and system for reducing RAID parity computation following a RAID subsystem failure. Ranges of RAID stripes are assigned to bits in a bitmap that is stored on disk. When writes to the RAID are in progress, the bit associated with the range of stripes in the bitmap is set. When a failure occurs during the write process, the bitmap is analyzed on reboot to determine which ranges of stripes where in the process of being written, and the parity data for only those ranges of stripes is recomputed. Efficiency is increased by use of an in-memory write counter that tracks multiple writes to each stripe range. Using the write counter, the bitmap is written to disk only after each cycle of its associated bitmap bit being set to a value of 1 and then returning to zero. The invention may be installed, modified, and removed at will from a RAID array, and this may be accomplished while the system is in operation.