Nomad backup restoration lacks native verification of recovered application data consistency

Direct Answer

Nomad's integrated backup mechanism does not include native verification of recovered application data consistency. The restore process captures job specifications, allocations, and Raft metadata but excludes attached storage volumes and task-level data. There is no supported method within Nomad to programmatically compare pre-backup and post-restore job allocations, and the system does not automatically verify application data integrity or detect version schema mismatches during restore.

Confirmed Limitations

The core limitations are:

• Excluded data: Persistent volumes, task stdout/stderr, and runtime state are not captured in backups
• No built-in diff: Nomad lacks allocation-state comparison tools for post-restore validation
• No schema validation: Version incompatibilities between backup and restore clusters are not detected automatically

Recommended Verification Workflow

Implement an external validation process:

1. Pre-backup export: Run nomad alloc status <alloc_id> for all allocations and save the JSON output
2. Post-restore comparison: After restoration, export allocation states again and compare using diff or a structured comparison tool
3. Persistent volume validation: Use filesystem checksums (e.g., sha256sum) on volumes before backup and after restore
4. Job spec validation: Compare HCL/JSON job specifications to ensure semantic equivalence

Version Mismatch Safeguards

To detect version schema mismatches:

• Record Nomad version during backup using nomad version
• Verify version compatibility before restore using HashiCorp's upgrade matrix
• Test restore procedures in a staging environment matching production versions

Conclusion

Operators must implement external checksum validation and manual inspection workflows. Nomad provides no native post-restore validation mechanisms for application data integrity.