Recon: Verifying File System Consistency at Runtime

File system bugs that corrupt file system metadata on disk are insidious. Existing file-system reliability methods, such as checksums, redundancy, or transactional updates, merely ensure that the corruption is reliably preserved. The typical workarounds, based on using backups or re-pairing the file system, are painfully slow. Worse, the re-covery is performed long after the original error occurred and thus may result in further corruption and data loss. We present a system called Recon that protects file sys-tem metadata from buggy file system operations. Our ap-proach leverages modern file systems that provide crash consistency using transactional updates. We define declar-ative statements called consistency invariants for a file system. These invariants must be satisfied by each trans-action being committed to disk to preserve file system in-tegrity. Recon checks these invariants at commit, thereby minimizing the damage caused by buggy file systems. The major challenges to this approach are specifying invariants and interpreting file system behavior correctly without relying on the file system code. Recon provides a framework for file-system specific metadata interpreta-tion and invariant checking. We show the feasibility of interpreting metadata and writing consistency invariants for the Linux ext3 file system using this framework. Re-con can detect random as well as targeted file-system cor-ruption at runtime as effectively as the offline e2fsck file-system checker, with low overhead.