prevent corrupt downloads after app and/or system crash #1644
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Otherwise, some pieces may be marked as finished in the control file though they have not yet been written to the storage file. This should prevent data loss and corruption when resuming downloads after an app crash. Signed-off-by: Ali MJ Al-Nasrawy <alimjalnasrawy@gmail.com>
This ensures that pieces are physically written to disk before marking them as finished in the control file. This should prevent data loss and corruption when resuming downloads after a system crash. Signed-off-by: Ali MJ Al-Nasrawy <alimjalnasrawy@gmail.com>
|
Hi @tatsuhiro-t, can you trrigger the travis-ci build again? It failed for OSX with a message I tested both linux and windows builds locally before submitting the PR and it worked as expected. |
|
Good catch. Thank you for PR. Merged now. |
|
linlong |
888520ll19811209ll
approved these changes
Dec 23, 2020
badwolf8u2
reviewed
Mar 27, 2023
| @@ -685,7 +685,7 @@ std::shared_ptr<DiskAdaptor> DefaultPieceStorage::getDiskAdaptor() | |||
|
|
|||
| WrDiskCache* DefaultPieceStorage::getWrDiskCache() { return wrDiskCache_; } | |||
|
|
|||
| void DefaultPieceStorage::flushWrDiskCacheEntry() | |||
| void DefaultPieceStorage::flushWrDiskCacheEntry(bool releaseEntries) | |||
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Problem:
Segments (technically , chunks of in-flight pieces) are marked as finished in the control file (*.aria2) as they are downloaded, though they might not be permanently written, yet, to the storage medium. In the case of app or system crash, and after resuming the download, some segments might have been lost, but the control file marks them as finished.
There is as significant chance for this to happen because of the delay introduced by the in-app disk cache and the OS buffers before the actual physical disk write. By default, auto-save interval is 60 seconds, while, in my case, linux buffer cache had >15s delay and the delay of in-app caches are highly dependent on the download speed because they are flushed once a whole piece completes, which makes it worse for slow network or parallel downloads.
Solution:
Flush data from the caches before saving control file.