some info about the hardware / node?
in short, defrag with ultradefrag and increase the timeouts via config.yaml