-
Notifications
You must be signed in to change notification settings - Fork 416
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
Implement delete_objs in fs and s3 storage backends. #395
Conversation
rust/src/storage/file/mod.rs
Outdated
@@ -134,6 +134,17 @@ impl StorageBackend for FileStorageBackend { | |||
async fn delete_obj(&self, path: &str) -> Result<(), StorageError> { | |||
fs::remove_file(path).await.map_err(StorageError::from) | |||
} | |||
|
|||
async fn delete_objs(&self, paths: &[String]) -> Result<(), StorageError> { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think paths: &[&str]
would be a slightly better signature.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I originally used &[&str]
but later met some problem when calling this method from vacuum
... It said that: temporary value created here. So I think maybe we should use String
instead of &str
here?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Ideally, I think the method should be defined as a generic function taking impl Iterator<AsRef<str>>
as input. But this trait is being used as trait object, so we can't use generic here, which makes things really annoying. I also think &[&str]
would have been a cleaner interface, but it will force us to allocate both Vec<String>
and Vec<&str>
in vacuum method, which has performance overheads :( I also don't really have a good suggestion on how to workaround this issue here.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@Dandandan Do you have any suggestion on this? Or we could go ahead with &[String]
for now?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think it's fine 👍
I added some comments @zijie0 |
@Dandandan Thank you for your suggestion, learned a lot from your PR :) |
* Batch delete in chunk. * Add default implementation for `delete_objs`.
I think this is a great improvement. I am wondering a bit how errors in the S3 implementation are handled, as the But I don't think that should be covered in this PR. |
Thank you both for this new optimization @zijie0 @Dandandan ! |
Description
Implement delete_objs in fs and s3 storage backends. Add corresponding test cases.
Related Issue(s)
Documentation