Current Answer: Yes, RAID Recovery Exists—But This 2024 Failure Happened While Replacement Was Disabled
The failed test is historically real. In ZimaOS 1.2.1, IceWhale explicitly said RAID replacement had been disabled because of a verification issue. The empty recovery dialog was therefore not proof that RAID recovery could never work; it was a version-specific disabled feature.
Do Not Recreate or Format a RAID When the Data Still Matters
Current recovery practice is much more conservative. If an existing array becomes unrecognized after reinstall or storage-database loss, preserve the member disks and recover the stored RAID configuration before creating a new array. The ZimaOS RAID recovery explains the current local-storage.db method and destructive fallback boundary.
The Linux mdadm array inspection is the upstream reference for identifying an md array.
A Failed Replacement Test and a Failed Update Were Two Different Problems
The thread later included a 1.2.0 system that could not see the update. That case turned out to be firewall-related: the update download was blocked. Do not mix an unavailable software update with the RAID member-replacement failure just because both appeared in the same discussion.
Test Recovery With Data You Can Lose
The original instinct was good: verify degraded mode and replacement behavior before trusting important data to an array. Use disposable test files, remove one member cleanly, confirm the array becomes degraded/read-only as expected, add a known replacement disk, and monitor rebuild state.
The kernel's Linux MD behavior provides the lower-level model.
RAID Recovery Is Not Backup
A successful RAID rebuild protects against a member-disk failure. It does not protect against accidental deletion, filesystem corruption, malware, controller mistakes or creating a new array over the wrong disks. Keep a second verified copy outside the array.
For current multi-drive hardware and recovery planning, the ZimaCube 2 storage gives the current platform context.
