Supporting internal consistency checking with consistency coded journal file entries
View Patent ↗Example systems, methods, and apparatus economize generating and processing incremental journal files while maintaining internal consistency. One example method determines whether a sequence number associated with a first inode description in a disaster recovery (DR) journal entry is out of sequence with a second corresponding inode description in a DR metadump. The example method controls a DR journal process to provide a file system inconsistency signal and to suspend application of the DR journal entry to the DR metadump. The suspending and signaling can occur upon determining that a first access time independent verification code computed from the first inode description does not match a second access time independent verification code computed from the second inode description.
1. A non-transitory computer-readable medium storing computer-executable instructions that when executed by a computer control the computer to perform a method, the method comprising:
upon determining that a sequence number associated with a first inode description in a disaster recovery (DR) journal entry is out of sequence with a second corresponding inode description in a DR metadump, where the sequence number associated is incremented when an item associated with the inode is accessed by a file system,
controlling a DR journal process to provide a file system inconsistency signal and to suspend application of the DR journal entry to the DR metadump as a function of determining that a first access time independent verification code computed from the first inode description does not match a second access time independent verification code computed from the second inode description; and
controlling the DR journal process to complete application of the DR journal entry to the DR metadump as a function of determining that the first access time independent verification code is consistent with the second access time independent verification code,
where the first access time independent verification code is computed in a file system transaction and is based on fields in a rollback object accessible to the transaction, where the fields in the rollback object are not affected by an access time only update, and
where the second access time independent verification code is computed from fields in the second corresponding mode, where the fields in the second corresponding mode are not affected by an access time only update.
2. The non-transitory computer-readable medium of claim 1 , the method including, upon determining that access to a file system entry will only update the access time for an mode associated with the file system entry, controlling the DR journal process to not create a DR journal entry for the access to the file system entry.
3. The non-transitory computer-readable medium of claim 2 , the method including, upon determining that the access to the file system entry will update more than just the access time for the mode associated with the file system entry, controlling the DR journal process to create a DR journal entry for the access to the file system entry, where the DR journal entry includes an access time independent verification code.
4. The non-transitory computer-readable medium of claim 1 , the first mode description comprising an abstract description of file system metadata associated with a file system mode and an update to a file system entry associated with the file system mode.
5. The non-transitory computer-readable medium of claim 4 , the second corresponding mode description comprising an abstract description of file system metadata associated with the file system.
6. The non-transitory computer-readable medium of claim 1 , the DR metadump comprising an abstract description of file system metadata.
7. The non-transitory computer-readable medium of claim 1 , where the DR journal process is configured to bring the DR metadump up to date with the file system.
8. The non-transitory computer-readable medium of claim 1 , where the first access time independent verification code is one of, a checksum, and a cryptographic hash and where the second access time independent verification code is one of, a checksum, and a cryptographic hash.