Uh oh!
There was an error while loading. Please reload this page.
Delete chunks if the move on an upload failed - #22128
Conversation
Uh oh!
There was an error while loading. Please reload this page.
MorrisJobke
commented
Aug 7, 2020
Tests fail. |
MorrisJobke
commented
Aug 7, 2020
Also the part file is still in the users home directory. Shouldn't it be deleted as well (this hasn't changed at all between master and this branch, but I just noticed it) |
99f79d3 to
0e45c99CompareMorrisJobke
commented
Aug 10, 2020
Autosquashed and force pushed |
MorrisJobke
commented
Aug 11, 2020
|
juliusknorr
commented
Aug 12, 2020
Are you sure that those are not from previous uploads? Then they should be cleaned up after 24 hours iirc. |
Signed-off-by: Julius Härtl <jus@bitgrid.net>
Signed-off-by: Julius Härtl <jus@bitgrid.net>
0e45c99 to
cda5cfcComparejuliusknorr
commented
Aug 12, 2020
Hm i assume this is also related.
I'll have another look |
Signed-off-by: Julius Härtl <jus@bitgrid.net>
juliusknorr
commented
Aug 13, 2020
Should be aligned to the previous behavior now as https://github.com/nextcloud/3rdparty/blob/b0afba6d6508a1c85332cf8c61e90ad91b289ebc/sabre/dav/lib/DAV/CorePlugin.php#L653 was called before Let's see what CI thinks about that. |
🤖 beep boop beep 🤖 Here are the logs for the failed build: Status of 31762: failuresqliteShow full logacceptance-app-files
Show full log |
MorrisJobke
commented
Aug 13, 2020
@juliushaertl I guess we backport this down to stable17? |
MorrisJobke
commented
Aug 13, 2020
/backport to stable19 |
MorrisJobke
commented
Aug 13, 2020
/backport to stable18 |
MorrisJobke
commented
Aug 13, 2020
/backport to stable17 |
This should fix#16517 where chunks don't get cleaned up if the move operation after a chunked upload fails for some reason, e.g. because the file is blocked by files_accesscontrol.
Steps to reproduce: