Thanks!

AH2: I see, and it would also be unsupported/operator-error to attempt to
unregister a controller while it is still running.


On Fri, Apr 3, 2026 at 1:38 PM Kevin Wu <[email protected]> wrote:

> Hi Alyssa,
>
> Thanks for the questions.
>
> RE AH2: You could also unregister a controller via the `kafka-cluster
> unregister-controller` command I am proposing in this KIP. Invoking either
> CLI command being introduced in this KIP with the wrong controller id can
> result in an operator accidentally unregistering a controller which is
> still running (has not been decommissioned). In this case, the controller
> needs to re-register since it is still part of the kafka cluster.
>
> RE AH3: Let me fix this sentence. I mean "The intention for unregistration
> is for it to occur after the operator decommissions a controller node. This
> means the node is no longer part of the kafka cluster, and the controller
> process associated with this node is not expected to come back." The
> unregister CLI command will return something to the user based on the
> UnregsiterControllerResponse.
>
> Best,
> Kevin Wu
>
> On Fri, Apr 3, 2026 at 3:27 PM Alyssa Huang via dev <[email protected]>
> wrote:
>
> > Few more questions actually
> >
> > AH1
> > Thanks, I was mainly trying to understand if dynamic reconfiguration
> offers
> > an incomplete downscaling UX without this KIP. Seems like that's the
> case.
> > I also missed the scenario with observer controllers so thanks for
> > clarifying that part.
> >
> > AH2
> >
> > > The registration manager of an unregistered controller should attempt
> to
> > > re-register with the active controller. This is to prevent accidental
> > > unregistrations.
> >
> >  I'm not sure if this makes sense to me, what would be an accidental
> > unregistration? Is the only trigger for unregistration the new
> > remove-controller command you're adding?
> >
> > AH3
> >
> > > The intention for unregistration is for it to occur after the operator
> > > decommissions a node which will no longer be part of the kafka cluster,
> > and
> > > is not expected to return.
> >
> > Why isn't the command expected to return?
> >
> > Thanks!
> >
> > On Fri, Apr 3, 2026 at 12:15 PM Alyssa Huang <[email protected]>
> wrote:
> >
> > > Thanks for the KIP Kevin!
> > >
> > > AH1
> > >
> > >> This means a stale controller registration from a controller that no
> > >> longer exists can block feature upgrades.
> > >
> > >
> > > And dynamic reconfiguration (specifically RemoveVoter) can cause stale
> > > controller registrations? Meaning anyone who has used RemoveVoter to
> > > permanently downsize their controller quorum would eventually need to
> > > unregister the controller to unblock future feature upgrades?
> > >
> > > Best,
> > > Alyssa
> > >
> > > On Fri, Apr 3, 2026 at 11:30 AM Kevin Wu <[email protected]>
> wrote:
> > >
> > >> Hi all,
> > >>
> > >> I would like to start a discussion on KIP-1312: Support unregistering
> > >> controllers. Below is the KIP link.
> > >>
> > >>
> > >>
> >
> https://cwiki.apache.org/confluence/display/KAFKA/KIP-1312%3A+Support+unregistering+controllers
> > >>
> > >> Thanks,
> > >> Kevin Wu
> > >>
> > >
> >
>

Reply via email to