Github user kl0u commented on the issue:
https://github.com/apache/flink/pull/5922
This seems like a nice suggestion.
I will do that and letâs see how it reads afterwards.
> On May 2, 2018, at 12:21 PM, Tzu-Li Tai <notificati...@github.com> wrote:
>
> @tzulitai commented on this pull request.
>
> In docs/dev/stream/state/broadcast_state.md
<https://github.com/apache/flink/pull/5922#discussion_r185452174>:
>
> > +
> +* ToC
> +{:toc}
> +
> +[Working with State](state.html) described operator state which is
either **evenly** distributed among the parallel
> +tasks of an operator, or state which **upon restore**, its partial
(task) states are **unioned** and the whole state is
> +used to initialize the restored parallel tasks.
> +
> +A third type of supported *operator state* is the *Broadcast State*.
Broadcast state was introduced to support use-cases
> +where some data coming from one stream is required to be broadcasted to
all downstream tasks, where it is stored locally
> +and is used to process all incoming elements on the other stream. As an
example where broadcast state can emerge as a
> +natural fit, one can imagine a low-throughput stream containing a set of
rules which we want to evaluate against all
> +elements coming from another stream. Having the above type of use-cases
in mind, broadcast state differs from the rest
> +of operator states in that:
> + 1. it has a map format,
> + 2. it is only available to streams whose elements are *broadcasted*,
> How about,
> "it is only available to operators which have a broadcasted input stream"?
>
> This might only be a matter of personal preference, so please take this
as a grain of salt.
> It's just that I somehow find it easier to understand when thinking in
terms of states and operators.
>
> â
> You are receiving this because you authored the thread.
> Reply to this email directly, view it on GitHub
<https://github.com/apache/flink/pull/5922#discussion_r185452174>, or mute the
thread
<https://github.com/notifications/unsubscribe-auth/ACS1qBiVh_nUWs6pz4XUSVk0w1mxN78Rks5tuYisgaJpZM4TlZsp>.
>
---
