macOS Tahoe 26.6.2 snapshot purge failure sits unfixed on a quiet Sunday (August 30, 2026)

Chapters
TL;DR
No new macOS defect was published or acknowledged on Sunday, August 30, 2026, and the Mac problem still live that weekend was the macOS Tahoe 26.6.2 snapshot purge failure diagnosed by The Eclectic Light Company on August 27, 2026. Time Machine's CacheDelete provider reported 0 bytes of purgeable space when about 94.9 GB should have been purgeable, so the deleted daemon refuses a copy onto a nearly full volume. As of August 30, 2026 Apple had published nothing about it and no macOS release had taken it away, so the only remedy is deleting local snapshots by hand in Disk Utility, which destroys your recent local restore points.
- August 30, 2026 was a Sunday with no macOS defect reporting anywhere I checked, so this post says so rather than inflating something.
- The live problem is unchanged since August 27, 2026: macOS Tahoe 26.6.2 refuses copies for lack of space while local snapshots hold tens of gigabytes.
- Time Machine CacheDelete answered the
deleteddaemon with 0 bytes when roughly 94.9 GB was purgeable. - One test volume disclosed its purgeable space 24 minutes after deletion; a second had disclosed none after 36 minutes.
- The workaround is deleting local snapshots yourself in Disk Utility, at the cost of your recent local restore points.
What was published about macOS on August 30
Sunday, August 30, 2026 produced no macOS defect reporting. 9to5Mac published five stories that day, on a CarPlay weather app, the next iPad mini, an Apple Pencil for iPhone Ultra, an Apple Watch Series 12 leak and weekend deals. No macOS bug, no Tahoe update, no macOS 26.7 item, no Golden Gate item and no security story appears among them.
The one Mac-technical piece that day, The Eclectic Light Company on August 30, 2026, names no bug, no version and no build number. Its subject is memory and storage supply: that "demand has outstripped supply" for high-speed memory and SSD NAND, that buyers pay around 25 pounds per GB of memory and 500 pounds per TB of SSD, and that a high-end Mac Studio M5 Ultra takes 10 to 12 weeks to arrive. That is a pricing story, not a fault.
So this is not an outage report, and I am not going to dress it up as one. It is an account of where the one actively discussed macOS defect stood on the day, and what you can still do about it.
The defect that was still live on macOS Tahoe 26.6.2
Before macOS allows a copy onto a nearly full volume, the deleted daemon asks every registered cache service, through the subsystem com.apple.cache_delete, how much space it can release. Time Machine's CacheDelete provider answers wrongly. In testing published on August 27, 2026 it returned 0 bytes "instead of the 94.9 GB that should have been purgeable", so the copy was refused. I covered that diagnosis in the August 27 post.
The symptom was described three days earlier, on August 24, 2026, in the preceding instalment: the Finder states "there isn't sufficient disk space" for a 12.89 GB copy onto a volume whose used space is in fact purgeable, tested on macOS Tahoe 26.6.2 in a 100 GB APFS container. At that point the cause was unknown.
Even when the space is eventually released, it arrives late and unpredictably. Two identically configured test volumes, each with over 90 GB purgeable, behaved differently: one disclosed its purgeable space 24 minutes after the files were deleted, and "the other still hadn't disclosed any purgeable space after 36 minutes had elapsed". That is why two Macs on the same release, in the same state, can disagree about whether a copy will fit.
The measurements, in one place
These are the published figures as they stood on August 30, 2026. None of them moved that day, because nothing new was published.
I have left out any claim about how many Macs are affected. Nobody has published a number, and the fact that two identical test volumes behaved differently means a count would be a guess.
| Measurement | Figure | Source date |
|---|---|---|
| macOS release tested | macOS Tahoe 26.6.2 | August 27, 2026 |
| Purgeable space Time Machine CacheDelete reported | 0 bytes | August 27, 2026 |
| Purgeable space that should have been reported | about 94.9 GB | August 27, 2026 |
| Test volume 1, delay before space appeared | 24 minutes | August 27, 2026 |
| Test volume 2, delay before space appeared | none after 36 minutes | August 27, 2026 |
| Copy the Finder refused in the earlier test | 12.89 GB | August 24, 2026 |
What Apple had acknowledged as of August 30, 2026
Nothing. Apple had made no acknowledgement of this behaviour, and no macOS release had removed it. The diagnosis that names Time Machine's CacheDelete provider is an independent researcher's work, not Apple's.
I do not know whether Apple has it logged internally, and I will not guess. The published record on this date is a named failing component, a reproducible symptom, no vendor statement and no fix.
What you can do today
1. If a copy or an install is refused for lack of space on macOS Tahoe 26.6.2, stop waiting for the automatic purge. The August 27, 2026 advice is exact: "If you need to free up space in a container, don't rely on the automatic purging of snapshots. Delete them yourself in Disk Utility."
2. Confirm your Time Machine backup disk is current before you delete anything. Local snapshots are your recent rollback points, so removing them is the cost of the workaround, and it is not a cost worth paying blind.
3. Open Disk Utility, select the volume, and delete the local snapshots you no longer need. Then try the copy again.
4. Check periodically for orphaned snapshots that are still present more than 24 hours after they were taken. The workaround is manual and has to be repeated, which is the real expense here.
5. If you are on macOS Tahoe 26.6.2 and copies are working, do nothing. There was no patch to install for this as of August 30, 2026, and deleting snapshots you still want buys you nothing.
Frequently asked questions
Is there a fix I can install for the snapshot purge failure?
Not as of August 30, 2026. No macOS release had addressed it and Apple had published nothing about it, so manual deletion in Disk Utility was the whole of the remedy on that date.
Why do two Macs on macOS Tahoe 26.6.2 behave differently?
Because the failure is not deterministic. Two identically configured test volumes with over 90 GB purgeable behaved differently on August 27, 2026: one released the space after 24 minutes, the other had released none after 36 minutes.