Pegasus R4i — APFS container corrupt, no valid checkpoint
I have a Pegasus R4i MPX module in a 2019 Mac Pro. The RAID array shows healthy in Promise Utility and macOS can see the physical disk (32TB), but the APFS container is corrupt and the volume won't mount.
Running fsck_apfs -n returns:
nx_block_count is 7813423094, while device block count is 7813417974
error: object (oid 0x1): o_cksum is invalid for object
error: no valid checkpoint
All 300+ APFS checkpoints disagree with block 0, so there's nothing for fsck to roll back to.
The block count mismatch is interesting — the container thinks it's ~20MB larger than the actual device. This makes me wonder whether a RAID event or firmware update at some point caused the LUN geometry to shift slightly, which then corrupted the superblock when APFS next wrote to it.
A few questions for anyone who's seen this:
- Is there a known issue with the R4i where a RAID rebuild or firmware update can cause a LUN geometry change that results in this kind of APFS superblock corruption?
- Does Promise Utility have any low-level diagnostic capability relevant here?
- Any R4i-specific recovery path I may have missed before I resort to an erase?
Thanks
