In my case one WDC20EADS drive just died.
I made the mistake to mess with the volume and RAID settings in the hope the "read & write" errors the 419P reported were just an incident. I tried to rebuild the RAID5 array with the ejected drive, but no luck. Kept getting read/write errors.
I ran a bad block scan and it reports more than 100 bad blocks (sectors), so my conclusion is that the drive has gone "bad". Funny enough the SMART data suggests there's nothing wrong with the drive.
I have RMA'ed the drive to Western Digital. I replaced the drive with a new WDC20EARS (64 MB cache). In the shop I heard that the "EADS" series has become "end of life" (= is not manufactured anymore). The main difference is cache size, which is irrelevant for a NAS. I have to add that WDC's website suggests the EADS series is still in production.
I should just have swapped the drive and let the NAS rebuild its RAID5 array, now I have to copy all data back again.. sigh