Jim Brunner

I'd like to propose that we take all unsorted sets and sort them. Clearly a sorted set is less messy than an unsorted one.

I'd like to add that we should do some explicit naming changes. In the old days... There was the current command, and then sometimes additional things which needed to be...

> IMO we already have existing generic mechanisms ... Agreed. The existing COB mechanism should be sufficient. Also agree that the client should be closed and the command should exit...

I think this idea may be misguided. Swap is never the problem. Swap is a SOLUTION to the problem of insufficient memory. Swap provides a safety net for transient OOM...

Can someone more clearly lay out the scenario here? I'm thinking that this is a case of: 1) creating a safe iterator on the hashtable, 2) deleting or moving the...

I've done a review on the iterator, and these are the things that need to be addressed. There are a few potential issues with long-lived (safe) iterators. One use case...

One other potential issue is with defragmentation. If the hashtable is moved during defragmentation, any active safe iterators will access the deleted memory. I don't think this can happen with...

We probably want a reasonably small size on the embedded key (and embedded value). Since jemalloc size buckets get progressively larger, the wasted space (internal fragmentation) will get (on average)...

First comment - 37 changed files??? Dang!

# Problem statement Valkey uses the operating system’s “fork” capability for: * save - creates a point-in-time snapshot, stored to an RDB file * full sync - attaches a replica...