Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Errors encountered during chmod operation causes runaway retry loop #3865

Closed
dhowland opened this issue Jan 2, 2017 · 4 comments
Closed
Labels
frozen-due-to-age Issues closed and untouched for a long time, together with being locked for discussion

Comments

@dhowland
Copy link

dhowland commented Jan 2, 2017

As reported on the forms, here, and in ticket #2727, certain configurations of ACL on FreeBSD/FreeNAS will cause a chmod to fail during file sync. When this happens, errors like these are seen on the console:

Puller (folder "xxxxx-xxxxx", file "Photo/IMAG0594.jpg"): dst create chmod: chmod /mnt/Documents/user/Desktop/Photo/.syncthing.IMAG0594.jpg.tmp: operation not permitted

These errors are repeated over and over for all affected files, during which processor utilization is observed to be 100% of a single core.

The sync state in the webGUI will appear to be stuck at whatever proportion of the dataset is affected. In my case this was 0%. On the NAS, all files were in place, however they were all renamed to .syncthing.*.tmp.

This issue popped up in the past as #1404. At that time, a workaround was added to use "Ignore permissions". That workaround does work to alleviate the ACL issue.

However, syncthing should abort the sync and report an error message to the user when it fails the chmod, similar to how it acts with other permissions issues (populating the "Failed Items" field and reporting the "operation not permitted" on each file). It should not go into a runaway retry loop.

@calmh
Copy link
Member

calmh commented Jan 2, 2017

I think the behavior is exactly the same as with other permission issues? It might look runaway because each operation is retried several times, for all affected files, every minute.

@dhowland
Copy link
Author

dhowland commented Jan 2, 2017

But, for example, if syncthing doesn't have write access to the shared directory, it will populate the "Failed Items" field with the errors for each file. In this case it never gives any indication that there is a problem. It just sits at "Syncing (0% )"

I let it go for 45 minutes, it never stopped grinding. After setting Ignore perms and starting over, it synchronized in about 2 minutes. If it's not technically a runaway loop, the symptom is close enough in my book.

@AudriusButkevicius
Copy link
Member

We it prints a message telling you to check the log whi hexplains the failure.
The loop is definately not a runaway loopas it gives up for a minute after 10 attempts for every file.

@dhowland
Copy link
Author

dhowland commented Jan 3, 2017

We it prints a message telling you to check the log whi hexplains the failure.

I never saw anything like this.

@calmh calmh closed this as completed Feb 7, 2017
@st-review st-review added the frozen-due-to-age Issues closed and untouched for a long time, together with being locked for discussion label Feb 8, 2018
@syncthing syncthing locked and limited conversation to collaborators Feb 8, 2018
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
frozen-due-to-age Issues closed and untouched for a long time, together with being locked for discussion
Projects
None yet
Development

No branches or pull requests

4 participants