You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@flink.apache.org by Cristian Lorenzetto <cr...@gmail.com> on 2019/08/14 10:01:25 UTC

why is not possible to handle a custom resource manager ?

i want develop a project using flink stack in a project using a custom
distributed system, so id like use my distrubuted system as resource
manager instead to overload the project with many other additional sockets
and code.

Is there a way for embedding flink project in my server without using other
external resources manager. For example a way for setting a
ResourceManagerInterface i dont know....

Re: why is not possible to handle a custom resource manager ?

Posted by Xintong Song <to...@gmail.com>.
Hi Cristian,

For the latest off-the-shelf Flink releases, I agree with Yang that running
a standalone cluster on top of your custom distributed system is AFAIK the
only way.

Customizing Flink and implement your own Flink ResourceManager is also an
option. Flink ResourceManager is a component that handles interactions with
external distributed systems, and we implemented differently for
standalone, Yarn, Mesos, and potentially K8s in future. Unfortunately, this
is not as easy as just implementing a set of interfaces at the moment,
and may need some knowledge about Flink deployment. In general, you need to
also touch the client for submitting jobs to your custom distributed
systems, cluster entry point to bring the Flink cluster up on your system
and specify using your customized ResourceManager, and task manager runner
to bring up task managers on your custom system. You can refer to the
flink-yarn module in the source codes for details.

Thank you~

Xintong Song



On Wed, Aug 14, 2019 at 1:43 PM Yang Wang <da...@gmail.com> wrote:

> Hi Cristian,
>
> I think you could have a try on standalone cluster on X(underlying cluster
> framework). By now it works well for kubernetes[1].
> If your custom distributed system support to start an application based on
> yaml/json configuration, then it will be very simple to start a session
> cluster. For job cluster, you need to build your own image with user jar
> and all dependencies included and then set the command of jobmanager to
> standalone-job.sh.
>
>
> [1]
>
> https://ci.apache.org/projects/flink/flink-docs-release-1.8/ops/deployment/kubernetes.html
>
> Best,
> Yang
>
> Cristian Lorenzetto <cr...@gmail.com> 于2019年8月14日周三
> 下午6:01写道:
>
> > i want develop a project using flink stack in a project using a custom
> > distributed system, so id like use my distrubuted system as resource
> > manager instead to overload the project with many other additional
> sockets
> > and code.
> >
> > Is there a way for embedding flink project in my server without using
> other
> > external resources manager. For example a way for setting a
> > ResourceManagerInterface i dont know....
> >
>

Re: why is not possible to handle a custom resource manager ?

Posted by Yang Wang <da...@gmail.com>.
Hi Cristian,

I think you could have a try on standalone cluster on X(underlying cluster
framework). By now it works well for kubernetes[1].
If your custom distributed system support to start an application based on
yaml/json configuration, then it will be very simple to start a session
cluster. For job cluster, you need to build your own image with user jar
and all dependencies included and then set the command of jobmanager to
standalone-job.sh.


[1]
https://ci.apache.org/projects/flink/flink-docs-release-1.8/ops/deployment/kubernetes.html

Best,
Yang

Cristian Lorenzetto <cr...@gmail.com> 于2019年8月14日周三 下午6:01写道:

> i want develop a project using flink stack in a project using a custom
> distributed system, so id like use my distrubuted system as resource
> manager instead to overload the project with many other additional sockets
> and code.
>
> Is there a way for embedding flink project in my server without using other
> external resources manager. For example a way for setting a
> ResourceManagerInterface i dont know....
>