Add concurrent access flag to the PlexusIoProxyResourceCollection #1
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
In practice it turns out that the PlexusIoProxyResourceCollection implementations are accessed by multiple threads but some of them (for example PlexusIoTarFileResourceCollection part of plexus-archiver) are not thread-safe. Add flag that indicates if the PlexusIoTarFileResourceCollection implementation supports concurrent access or not.
|
Great stuff, thanks a lot ! |
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Outdated suggestions cannot be applied.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.
Plexus Archiver added support for concurrent zip creation that does lead to a regression - adding the content of tar archive to a zip file is no longer possible as the tar archive is not thread safe. There are two bug reports about that: codehaus-plexus/plexus-archiver#10 and MASSEMBLY-789
There are two ways to solve this - either the tar code must be made thread safe or the zip archiver should be able to add some of the entries in non-concurrent way.
Tar archives does not have indexes for entries. And even if they did (one could be build while iterating over the entries) you can't just skip to the location of the entry as usually the tar are archived. You must decompress the tar file (at least until the entry) for every entry. Another possibility is to unpack the tar archive to a temporary location and then build the index while iterating over the entries. This index could be used to access directly the entry of the unpacked tar file.
Both of this solutions require a lot of unnecessary CPU and/or IO usage so I think the end result will be worse performance even with the concurrent zip utilization of all available cores.
I think the better solution is to acknowledge that there are some resource collections that should be accessed in non-concurrent way so the concurrent jar creator (or any other code that works with such collections) access them in single thread manner. The big question is how. With this pull request I'm proposing to add a new method to the
PlexusIoProxyResourceCollectioninterface. What do you think?I also considered adding new marker interface. But there is heavy usage of proxies for resources and resource collections and I think even if implemented would not result in good and easy to maintain code.