Sorry to join so late, but did we consider a single set of task-count
metrics and using tags to scope each data point to a specific
connector and worker (and in the future perhaps also user)?

It will make analysis of the data easier - someone may want to
breakdown tasks by both worker and connector to detect imbalanced
assignments.

Are there downsides to this approach?

And a small nit: it will be good to capture in the KIP what are the
expectations regarding overlap and disjointness of the proposed
metrics. For example, is running+paused+failed = total? Can a task be
failed and destroyed and therefore count in 2 of those metrics?

On Thu, Jun 6, 2019 at 12:29 PM Cyrus Vafadari <cy...@confluent.io> wrote:
>
> Konstantine,
>
> This is a good suggestion. Since the suggestion to add 2 additional
> statuses analogous to the 3 proposed, it is a very minor change of no
> structural consequence to the KIP.
>
> I've updated the KIP to incorporate your suggestion, and any voters who
> disagree should definitely respond in the thread.
>
> Cyrus
>
> On Thu, Jun 6, 2019 at 11:16 AM Konstantine Karantasis <
> konstant...@confluent.io> wrote:
>
> > Thanks Cyrus,
> >
> > this is a nice and straightforward addition.
> >
> > I'm +1 too, but I'd like to return with a question here as well regarding
> > whether the unassigned tasks will be taken into account.
> > Especially after KIP-415 we might start seeing this status for specific
> > periods of time. Therefore, I think it's a meaningful addition.
> > Then there's the `destroyed` status which might be a lot more transient but
> > we could also include for the sake of completion.
> > Check org.apache.kafka.connect.runtime.AbstractStatus for the list of all
> > possible statuses.
> >
> > Konstantine
> >
> > On Wed, Jun 5, 2019 at 4:32 PM Randall Hauch <rha...@gmail.com> wrote:
> >
> > > Thanks, Cyrus.
> > >
> > > +1 (binding)
> > >
> > > Randall Hauch
> > >
> > > On Wed, Jun 5, 2019 at 10:36 AM Andrew Schofield <
> > > andrew_schofi...@live.com>
> > > wrote:
> > >
> > > > +1 (non-binding)
> > > >
> > > > Andrew Schofield
> > > >
> > > > On 05/06/2019, 14:04, "Ryanne Dolan" <ryannedo...@gmail.com> wrote:
> > > >
> > > >     +1 (non-binding)
> > > >
> > > >     Thanks
> > > >     Ryanne
> > > >
> > > >     On Tue, Jun 4, 2019, 11:29 PM Cyrus Vafadari <cy...@confluent.io>
> > > > wrote:
> > > >
> > > >     > Hi all,
> > > >     >
> > > >     > Like like to start voting in the following KIP:
> > > >     >
> > > >     >
> > > >
> > >
> > https://nam01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fcwiki.apache.org%2Fconfluence%2Fdisplay%2FKAFKA%2FKIP-475%253A%2BNew%2BMetric%2Bto%2BMeasure%2BNumber%2Bof%2BTasks%2Bon%2Ba%2BConnector&amp;data=02%7C01%7C%7C95f8a8ebb4a44882773808d6e9b65983%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C636953366722392496&amp;sdata=vbE%2BjrAapcQ68Vnwh5OkY1FFoOzFHs9rZRaPHlwqxSU%3D&amp;reserved=0
> > > >     >
> > > >     > Discussion thread:
> > > >     >
> > > >     >
> > > >
> > >
> > https://nam01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.apache.org%2Fthread.html%2Fbf7c92224aa798336c14d7e96ec8f2e3406c61879ec381a50652acfe%40%253Cdev.kafka.apache.org%253E&amp;data=02%7C01%7C%7C95f8a8ebb4a44882773808d6e9b65983%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C636953366722402501&amp;sdata=0JpQuCpTKwJyOjWH8cM%2B6eU%2FjNT28eE7xvMOBQgghjA%3D&amp;reserved=0
> > > >     >
> > > >     > Thanks!
> > > >     >
> > > >     > Cyrus
> > > >     >
> > > >
> > > >
> > > >
> > >
> >



-- 
Gwen Shapira
Product Manager | Confluent
650.450.2760 | @gwenshap
Follow us: Twitter | blog

Reply via email to