Okay. Doesn’t look like a file system issue. Has your node been online for a while, or has it recently been restarted? When the node is first started it runs through the directories to recalculate used space. This can take some time as littleskunk said. Is there high disk activity? If disk activity is low, and you have had the node running for more than 24 hours, you could try the steps outlined in this post:
You will need sqlite3 installed in order to recreate the piece_spaced_used.db database. This will force the node to recalculate space used.