[
https://issues.apache.org/jira/browse/FLINK-3299?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15143496#comment-15143496
]
Stephan Ewen commented on FLINK-3299:
-------------------------------------
One thing interesting about save points is that one can use them to "fork" off
new versions of the application:
Take a save point and restart a second copy of the streaming job (slightly
modified, think A/B testing).
In that case, the savepoint is explicitly not tied to an application.
The restriction of that is that the savepoint is completely self-contained, so
it does not depend on data in another system (would work for FileSystem,
RocksDB, TFiles, MapFiles, ...)
Some of these ideas are outlined here:
http://data-artisans.com/how-apache-flink-enables-new-streaming-applications/
> Remove ApplicationID from Environment
> -------------------------------------
>
> Key: FLINK-3299
> URL: https://issues.apache.org/jira/browse/FLINK-3299
> Project: Flink
> Issue Type: Improvement
> Reporter: Ufuk Celebi
> Assignee: Ufuk Celebi
> Fix For: 1.0.0
>
>
> {{ApplicationID}} is used to identify an application across many job
> submissions (for example after restoring from a savepoint). This is currently
> exposed in the {{Environment}}, which might be unnecessary.
> State backends, which need the ID can generate it themselves and store it as
> part of their state handle.
> This has to be checked with the DB state backend, which currently uses this.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
