Binbin
> Another option is to reject the config change during rewrite? yes, this also make sense, the two options work for me, there is another discusstion in #887
> Another option is to reject the config change during rewrite? This may be a breaking change in some ways, @zuiderkwast i am going to keep this one, do you...
yes. It is a small bug
we are saving nodes.conf in beforeSleep, and event save it without fsync, there is a time gap i think, and shutdown has its own save seems good to me (it...
> Maybe we can extend this more common, for something like KEYS h[ae]llo , we can translate it to exactly match as hallo and hello it seems like a good...
ok, i did not read https://github.com/redis/redis/pull/12536 at that time so i missed this context, sorry about that, and sorry i am not realizing it > We implemented this feature on...
oops, sorry, i just merged #946, and did't see this PR before. Different people handled different PRs at different times...
Agree with this, i want to do that. Another thing i want to deal with is, if we pass a large numreplicas, can we implicitly consider it as WAIT ALL?...
> Maybe it should be made clear that we will only wait for the "ACTIVE" replicas which are in the process of steady state syncing or psync? Yeah, i think...
this seem like a interesting feature, did not review, just drop a comment that approve the concept.