WIP: Codegen-free FASTER C# #67
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
badrishc
added
enhancement
…, for in-memory operations. BasicFASTERTests test cases now pass. There is no performance regression for 50:50 read/upsert YCSB workload. ManagedSample1 is now based on generic FASTER. Support for non-blittable objects has not yet been added. Codegen fails - but can be ignored for now.
…terface. Now, there are no interface restrictions or requirements on any of the FASTER key/value types.
|
Work is complete. All tests pass, with no performance regression. Merged, closing PR. |
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Outdated suggestions cannot be applied.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.
There have been some reports of code gen in FASTER C# causing slow instantiation (5 secs), plus the integration with arbitrary user types has overheads and restrictions (e.g., with access). Therefore, we are experimenting with a code gen free version of FASTER in this PR.
The basic idea is to use calls from System.Runtime.CompilerServices.Unsafe to enable in-place updates of .NET types, and manage the hybrid log without requiring a pointer based core API. I benchmarked Unsafe.AsRef and it appears we can do a billion of these on one core, per second. We are slowly updating parts of the FASTER code, with constant performance benchmarking, to ensure that changes we attempt do not regress performance significantly.
The current status of the PR is: I have remove Key* and Value*, and converted "Functions" to not be static. Performance has not degraded yet. In fact, changing callbacks to user Functions from static to instance-based, has given a 10% hike in performance at scale (72 cores), not sure why.
Comments appreciated - including from @mike-barnett @ReubenBond @gunaprsd.