Ashley Pittman

For this PR whilst it's still work-in-progress I suggest that you should modify https://github.com/daos-stack/daos/blob/master/src/tests/ftest/dfuse/daos_build.py#L206 to do a "git checkout lei/DAOS-14149" so that the is trying to build itself for the...

> > Looking at the test logs there are about 280k read calls to dfuse vs 2.8m read calls for the `nocache` version of the test, therefore this is intercepting...

https://github.com/daos-stack/daos/pull/14023 fixes the issue with misuse of `seekdir` in the test, after which it appears to pass on /tmp, dfuse and pil4dfs. Please review and merge into this PR if...

> > #14023 fixes the issue with misuse of `seekdir` in the test, after which it appears to pass on /tmp, dfuse and pil4dfs. Please review and merge into this...

Sorry, I was going to comment on this one. `fuse_session_custom_io` doesn't have the same protection for adding entries to the end of structs that fuse_session_new does so extending this struct...

I think the trick that `fuse_session_new` uses is a good one, what would be ideal would be if `fuse_session_custom_io` had used it from the start. I was putting `fuse_session_new` forward...

> It would break the API - the worst case. We can update `fuse_session_custom_io` That's probably a good idea, at least as a start. It looks like this affects 3.13-3.16...

This looks awesome, thanks for researching this.

Error numbers are already known at https://myenergi.info/could-myenergi-provide-a-list-of-error-numbers-ple-t1882.html The priority is also known and is diverter priority isn't it? I've not used it personally but have seen references to it before.

@tensiondriven thank you for reaching out again, from following your links it does seem like this is a genuine, reproducible problem which is the first step to getting something like...