Oran Agra

> Also, `kqueue` and `evport` are working with nanosecond precision as well. It's good to keep `epoll` in sync with the current `kqueue` and `evport` in `redis`. they're working with...

i agree with most of the above statements, if this API was added in kernel 2.6 we'd just make a switch, but in the current situation this does add complexity....

any_shard is the default (for commands that don't take key arguments) here's what the docs says (a few lines above the ones modified): > The default behavior a client should...

@uglide do the clients do that because that was their behavior before hints were available? or because they misread the docs? do you see any reason not to fix the...

When the db doesn't have a lot of data, and is repeatedly running short defrag cycles, these start and stop prints will flood the log. They must be throttled in...

in 1 we propose that HSET will be skipping both the hset and hdel KSN, right? i.e. for a field that didn't exist before, and add an hdel ksn for...

you mean if HSETEX is like an Lua script that calls HSET+HEXPIRE.. but it's not, it's a single command that does it atomically. from that perspective, the key was never...

>if the module thread is always the highest priority on a cpu, it will take a long time to let the main thread to reacquire the GIL. but the module...

@DvirDukhan @MeirShpilraien i remember some discussion about Redis Search around this topic.

@MeirShpilraien i remember an issue with RM_Yield and trying to give up the GIL. calling usleep separately wouldn't have helped (unless maybe you released the mutex before?)