Hi Michael,
Thanks for your feedback.
> What are the use cases where a client needs to know the cluster health
> other than for auto-failover?
For the time being, I only think of the case of auto-failover.
> The question of cluster health is
> complicated, given that Pulsar is a distributed sys
Thanks for your response Cong Zhao.
> I think don't need to handle it, because the client API should be consistent
> with the `HEALTH_CHECK` command result, and users can retry it if they need.
I think I may be coming at this from a different perspective. A client
application does not need to kn
Hi Michael,
> I think the current PIP might need some clarification on how errors
> are handled. For example, if a single broker fails to respond because
> it was being restarted, how would the client handle that kind of
> failure with this feature?
I think don't need to handle it, because the cl
Thanks for your replies Cong Zhao.
I think the current PIP might need some clarification on how errors
are handled. For example, if a single broker fails to respond because
it was being restarted, how would the client handle that kind of
failure with this feature?
> This is a good definition of c
Hi Michael,
Thanks for your feedback.
> I define a client's primary cluster as "healthy" when it is "healthy"
for all of its producers and consumers. I define a healthy producer as
one that can connect to a topic and publish messages within certain
latency and throughput thresholds (configured by
I'd like to clarify the motivation for this PIP. My understanding is
that the primary motivation is to give clients a robust way to
classify a cluster as "healthy". The initial beneficiary of this
feature is the auto failover use case. I think the feature makes
sense, but before using the broker's
Yes, there may have multiple clients request the HC at the same time in the
AutoFailover case, so we should add some cache to reduce broker load.
On 2022/06/22 12:55:49 Enrico Olivelli wrote:
> Il giorno mer 22 giu 2022 alle ore 14:45 Cong Zhao
> ha scritto:
> >
> > Hi Enrico,
> >
> > > Also, I
Il giorno mer 22 giu 2022 alle ore 14:45 Cong Zhao
ha scritto:
>
> Hi Enrico,
>
> > Also, I would like to understand in which usecase you can use the
> > binary endpoint and not the HTTP endpoint.
>
> We can't use the HTTP endpoint when the client did not have the admin auth to
> do a health chec
Hi Enrico,
> Also, I would like to understand in which usecase you can use the
> binary endpoint and not the HTTP endpoint.
We can't use the HTTP endpoint when the client did not have the admin auth to
do a health check. but we need it in some cases such as auto failover on the
client-side (htt
I believe that this proposal is too broad.
the PIP reads about:
- adding HEALTHCHECK to the binary protocol
- add a HEALTHCHECK cache on the broker
Also, I would like to understand in which usecase you can use the
binary endpoint and not the HTTP endpoint.
Health Check is good for scripts and fo
Hello Pulsar Community,
Here is a PIP to introduce the HEALTH_CHECK command in the binary protocol.
I look forward to your feedback.
PIP: https://github.com/apache/pulsar/issues/15859
Thanks,
Cong Zhao
11 matches
Mail list logo
