Tyler Neely

@cberner that's a good point, thank you for digging in! As time goes on, the incentive for not using O_DIRECT becomes smaller and smaller, and I'm starting to think that...

O_DIRECT has well-known pros and cons, and especially in combination with io_uring seems to be the clear winner for point-read situations like what tends to dominate b-tree like storage engines.

the initial implementation should borrow heavily from the ARC, using ghost lists on both an MRU & MFU cache to determine how perf would be if they were the only...

How large is the database? Is it larger than 4gb?

@SilentCicero there are a few factors that make this lower on the priority list for the time being: 1. likely 32-bit issue 2. involves a lot of code paths that...

I'm currently looking into this approach for handling this issue: https://github.com/spacejam/sled/issues/1093

@SaulDoesCode not inappropriate at all! And thank you for the kind words :D Transactions are a very high priority right now. Here's the status of the different pieces that need...

After months of thought around this, a few opportunities (constraints) have popped up: * we want to build more than just b-trees on top of the pagecache: relational tables, hash...

Initial sketches of versioning are underway in the [tyler_versions](https://github.com/spacejam/sled/tree/tyler_versions) branch. Basically, I've wanted to falsify a few competing designs first to make sure that the complexity of versioning pushed to...

~~pegging can also be implemented by just delaying the completion of the very first reservation made in the batch, so it's actually 0 overhead~~ it's a little trickier than that,...