Viktor Söderqvist
Viktor Söderqvist
> Fair. However, I don't see a mechanism for a user to disable printing the human nodename i.e. one can't set to empty. Rather setting it to empty would lead...
> Actually it won't. `CLUSTER SHARDS` shows the node's hostname but not the human nodename. Oh, right. But it's not unlikely that we'll add human nodename to CLUSTER SHARDS in...
> So the main purpose is for the local env debug? It is not a real request for the production env? @enjoy-binbin It can be very useful when reading logs...
> @zuiderkwast Should we merge this now that it's approved? Yes, but first I want to give some time for the other reviewers @enjoy-binbin and @hpatro to comment. I will...
@PingXie I updated the PR description. Please check if it's correct and edit again if it's not.
Two commits need sign off. They're called "Update src/valkey-cli.c".
I agree with keyword. Go for `TIMEOUT ms`.
@murphyjacob4 will you have some time to close on these details and get this merged?
You can use any command on the command line, e.g. `valkey-benchmark -r 1000 -n 1000000 SET key:__rand_int__ IFEQ value:__rand_int__ value:__rand_int__`. You can also contribute a predefined setifeq test to valkey-benchmark...
Using WATCH is indeed a problem in clients that are reusing the same connection for multiple lightweight threads or similar. I imagine WATCH is not handled well in for example...