I came across KIP-1161(https://cwiki.apache.org/confluence/x/HArXF), which made me think again if we need the warning. The kip changed the implicit failure caused by an empty value of bootstrap.servers to an explicit one without any intermediate warnings. I think it's a sensible thing to do in this KIP as well. what do you think?
On Thu, Dec 4, 2025 at 1:00 PM Nikita Shupletsov <[email protected]> wrote: > > Hi 안상혁 , > > sorry for a delayed response. > > Yeah, that sounds like a good idea. Could you please update the KIP to > reflect the changes we discussed? > > On Fri, Nov 28, 2025 at 6:30 PM 안상혁 <[email protected]> wrote: > > > > Hi Nikita! > > > > Thanks for pointing that out! > > I checked org.apache.kafka.streams.state.HostInfo, and you are right. (it > > already has logic to parse and validate the endpoint.) > > > > How about extracting that parsing/validation logic into a shared static > > utility? > > > > I'd prefer to avoid introducing a direct dependency on the HostInfo class > > within StreamsConfig and ConfigDef, as HostInfo belongs to the state layer. > > So instead of using HostInfo directly, we could: > > > > Extract the logic currently inside HostInfo into a static utility method > > (or make it accessible statically). > > > > Have both HostInfo and the new StreamsConfig validator use this shared > > logic. > > > > > > This way, we ensure consistency between HostInfo and the config validation > > while keeping the dependencies clean. > > What do you think?
