perf: remove crawler scrape lock #425
Merged
Conversation
This file contains 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
This PR removes the async lock from the crawlers and replaces it with a semaphore with a capacity of 20 In reality, neither the lock nor the semaphore are needed. The requests are actually limited by the `request_limiter` of each crawler, not the lock. However, i could not remove the lock because it will brake the logic for the UI scrape queue, that's why i replaced it with a semaphore instead. The lock was making each crawler behave synchronously and the `request_limiter` was never close to being reached. This was only for the crawlers. The downloaders have a semaphore already with a different capacity per domain, so they were not affected. The default capacity for each downloader is 3, defined by `--max-simultaneous-downloads-per-domain` ## Disadvantages The only drawback of replacing the lock with a semaphore (or eventually removing the lock altogether) is that the `request_limiter` is defined per crawler and almost all crawlers right now have a generic `10 requests / sec` limit. Some crawlers may require fine tuning the limiter to make sure CDL does not trigger `429s`
NTFSvolume
added a commit
to NTFSvolume/CyberDropDownloader
that referenced
this pull request
Jan 5, 2025
Moved to PR jbsparrow#425
jbsparrow
approved these changes
Jan 5, 2025
NTFSvolume
added a commit
that referenced
this pull request
Jan 5, 2025
* fix: use a dataclass for reddit posts Should fix #426 * refactor: pass `scrape_item` as origin for `web_pager` (chevereto) * fix: "Loose Files" not being created (all crawlers) * refactor: add custom MediaFireError (mediafire) * fix: scrape error codes * refactor: make `RealDebridError` Inherit from `CDLBaseError` * refactor: move `RealDebridError` to `clients.errors` * refactor: move `VALIDATION_ERROR_FOOTER ` to `constants` * fix: undo crawler semaphore Moved to PR #425
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.
This PR removes the async lock from the crawlers and replaces it with a semaphore with a capacity of 20
In reality, neither the lock nor the semaphore are needed. The requests are actually limited by the
request_limiterof each crawler, not the lock. However, i could not remove the lock because it will brake the logic for the UI scrape queue, that's why i replaced it with a semaphore instead.The lock was making each crawler behave synchronously and the
request_limiterwas never close to being reached. This was only for the crawlers. The downloaders have a semaphore already with a different capacity per domain, so they were not affected. The default capacity for each downloader is 3, defined by--max-simultaneous-downloads-per-domainDisadvantages
The main drawback of replacing the lock with a semaphore (or eventually removing the lock altogether) is that the
request_limiteris defined per crawler and almost all crawlers right now have a generic10 requests / seclimit. Some crawlers may require fine tuning the limiter to make sure CDL does not trigger429sThe I/O to the requests cache will increase but that one is already overloaded in some cases. Needs a fix for the database lockups.