This comes from the fact that the QNap SNMP Daemon shows the number of blocks, and the block size is set to 4 - if the default block-size is 1024.

Use a multiplier on the MRTG Line to have the correct partition size displayed.
e.g.:

Target[nas_md0_data]: 1.3.6.1.2.1.25.2.3.1.5.33&1.3.6.1.2.1.25.2.3.1.6.33:[COMMUNITY]@[IP-Address] * 4000

This will correct the data output. Of course - if you can do that on the RRD tool display, it would be easier. BTW - if the data is coming in as counter - use 4096 - to adapt to byte-counter. 

 This can happen when the Disks won't spin up fast enough. Even if enough disks spin up - it can be quite bad, as it will run in degraded mode or mount the spare-drive into the Raid 5.

Best thing to do, is to login remotely into the SS839 device, and issue the rboot command. This will trigger a warm-boot, and most disks will be hopefully back up.