Christopher Berner

Ah yes, (3) is this bug: https://github.com/cberner/redb/pull/501 it happens in those tests and also if you reopen a database on MacOS with `open()` (doesn't happen with `create()`)

I think this will fix (2): https://github.com/cberner/redb/pull/502

Oh yes, I ran benchmarks. mmap is something like 20-50% faster for writes and 1.2x to 3x faster for reads depending on the number of concurrent threads. The biggest benefit...

Hmm, well that's definitely bad. @niftynei do you have any way of reproducing this? I could create a branch with more debugging info. Just from that backtrace it's hard to...

> it happened again pretty quickly while running again, so i'd say yes, seems fairly reproducible. i was just about halfway through indexing. > > i went ahead and started...

I just merged a fix for this, can you try with redb commit? https://github.com/cberner/redb/commit/1293d4fbee6d594bbc1324b38ffd5e6c8b66a9f8

There are no sources of randomness in redb. However, there are quite a few constraints if you want an exact byte for byte copy of the db (and I've never...

I think you'll need to replace the `HashMap` caches in ord with a data structure with deterministic order, like a `BtreeMap`. For example, this looks like it will cause non...

What error message do you get? There's this open issue in redb: https://github.com/cberner/redb/issues/519

I'd suggest patching ord to use `create()` / `open()` (https://docs.rs/redb/latest/redb/struct.Builder.html#method.create) instead of `create_mmapped()`. If that solves it, then this is most likely an issue with the mmap management in redb....