we have a CephFS with a replicated data pool and an additional EC data pool. All files store their data in the EC pool.

But "ceph df" shows that the replicated data pool contains over 200 million objects with a total size of 700 bytes. So the vast majority are empty objects.

How can this be?

I don't do cephfs, but at least radosgw likes to stash information into zero-sized objects, so the sum of the sizes of those objects was small, but the data usage would of course be larger than zero on those rgw pools. Perhaps cephfs does the same? "ceph df detail" on one of our rgw clusters show this: POOL ID PGS STORED (DATA) (OMAP) OBJECTS USED (DATA) (OMAP) %USED MAX AVAIL QUOTA OBJECTS QUOTA BYTES DIRTY USED COMPR UNDER COMPR default.rgw.buckets.index 7 512 51 GiB 0 B 51 GiB 9.57k 154 GiB 0 B 154 GiB 0.02 225 TiB N/A N/A N/A 0 B 0 B the "DATA" column is zero, even though there are 9.57k objects in there, but OMAP is lots more, probably meaning it stashed information into the metadata of said zero-sized objs. If we stat one of the .dir.xxxx objects, it says: rados stat -p default.rgw.buckets.index ".dir.b9865dc1-e648-4479-abcd-ed812345678.144664284.33737.1.20" default.rgw.buckets.index/.dir.b9865dc1-e648-4479-abcd-ed87654321.144664284.33737.1.20 mtime 2026-07-30T09:48:05.000000+0000, size 0 -- May the most significant bit of your life be positive.