neoremind commented on PR #16279:
URL: https://github.com/apache/lucene/pull/16279#issuecomment-6338121542
Hi @jimczi, I'd like to revive this PR. As time goes, I find this benchmark
useful for multiple other PRs, and it can also be seen as a Java version of
`fio`. I run each strategy (buffered io, pread, direct io, mmap with various
hint and batched prefetch) vs. `fio` doing the same job, and the results match,
it basically mirrors `fio` close and be value to vet Lucene's I/O primitives.
Given that, I think it's worth doing.
New revision:
- Direct I/O now uses Lucene's DirectIODirectory instead of raw FFI O_DIRECT.
- Some small cleanups.
- A more complete javadoc that explains what each strategy does, how to run
it, parameters, and what to watch out.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
