You have a RAID-1 with 2 disks and one disk fails. You change the failed disk and everything seems good as the disk starts recovery. But recovery stops as the old disk with data on it changes into state ECC-ERROR. So you have your data on a disk with status ECC-ERROR and a new disk with status DEGRADED which stopped recovery. To continue recovery you have to ignore the ECC-ERROR and tell the controller to rebuild the DEGRADED disk by the following command. In the example we manage unit 0 on controller 0 and disk p1 is degraded.
tw_cli /c0/u0 start rebuild disk=1 [ignoreECC]