From a quick review of the log, this doesn’t look like “Time Machine choosing to do full backups” so much as “Time Machine can’t reliably *see or update* the existing backup set on the network target, so it behaves like it has to start over (or never completes enough state to remain incremental).”
The biggest red flags are the Permission denied / Operation not permitted when reading volume capabilities for the mounted Time Machine share, plus “Resource busy” when writing com.apple.TimeMachine.MachineID.plist inside the sparsebundle. If Time Machine can’t read the destination’s capabilities consistently and can’t update its identity/plist files, it can’t cleanly maintain the incremental chain.
The other repeating error, tmutil … fs_snapshot_list failed: Operation not supported, strongly suggests the Mac isn’t able to enumerate local filesystem snapshots for the source volume at that moment (or the source isn’t presented as snapshot-capable). Modern Time Machine relies heavily on APFS snapshots to produce fast incrementals. If snapshot operations aren’t available/working, backups can become much larger and “full-like,” and if the network destination is also flaky/permission-blocked, you’ll see repeated resets.
The BACKUP_IN_PROGRESS_REQUEST_DROPPED (46) is consistent with a backup getting wedged and new requests being ignored/dropped.
Ok, with all that said, the most likely causes are:
- NAS / SMB share permissions or capability reporting are broken/inconsistent: (Permission denied, Operation not permitted, can’t read capabilities, can’t get remount key).
- Sparsebundle is being held open/locked: (Resource busy). Often caused by NAS-side indexing/AV, another Mac trying to back up to the same bundle, stale SMB handles, or a prior mount not cleanly detached.
- Source snapshot support is failing: (fs_snapshot_list … Operation not supported). These can lead to huge deltas and fragile incrementals.
- Mount path / share identity instability. If the share mounts differently each time (Bonjour name changes, different subpath under /Volumes/.timemachine/...), Time Machine can treat it as a “new” destination.
Finally, here are some steps you can take to hopefully resolve this:
- Confirm the destination is stable and writable:
- On the Mac: open Time Machine settings and verify it’s always pointing to the same network disk/share.
- Using the Finder, connect to the share manually, create/delete a test folder (confirms real write access).
- On the NAS: ensure the Time Machine user has full read/write on the Time Machine share and the share is explicitly configured for Time Machine over SMB.
- Eliminate Resource busy / locking:
- Ensure only one Mac is using that specific sparsebundle (each Mac should have its own bundle).
- Temporarily disable NAS-side indexing/antivirus on the Time Machine share (if enabled).
- Eject the Time Machine volume, reboot the NAS (or at least restart SMB), then reboot the Mac and try again.
- Check whether Time Machine is actually re-creating the backup set:
- On the Mac (Terminal):
- tmutil listbackups (if it returns nothing / resets each run, it’s not maintaining the chain)
- tmutil destinationinfo (confirms what destination TM thinks it’s using)
- Fix the snapshot issue on the Mac:
- Confirm the source volume is APFS and not an unusual setup.
- Make sure there’s adequate free space on the Mac (snapshots can fail when space is tight).
- If third-party security software is installed, temporarily disable it and retest (some tools interfere with backupd access patterns).