We have experienced yet another RAID battery failure in the same batch of these hadoop workers.

This is around the 5th node to fail in a similar way recently.

The impact of this is a reduction in disk write performance for each of the affected nodes, as the controller switches to write-through mode.