[jira] [Commented] (FLINK-15911) Flink does not work over NAT

["Xintong Song (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/FLINK-15911?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17041749#comment-17041749 ] 

Xintong Song commented on FLINK-15911:
--------------------------------------

I've spent some time looking into this issue, together with FLINK-13171 & FLINK-15154. I think these three ticket can be resolved together. Therefore, I'm closing the other two tickets and working on this ticket.

I'm posting my proposed solution to this issue below, in case someone has a different thought.

In addition to the current configuration options ('jobmanager.rpc.[address|port]', 'taskmanager.[host|rpc.port]'), I propose to introduce the 4 new config options: '[jobmanager|taskmanager].rpc.bind-[address|port]'.
- The original config options will be used to decided the exposed address and port, same as they are now.
- The new config options will be used to decide the bind address and port, without default values.
- If the bind address/port is not specified, by default JM/TM binds to the IP address 0.0.0.0 and the port that is exposed, same as the current behavior.
- Note that, while '[jobmanager|taskmanager].rpc.port' allows specifying a range of ports (e.g., 50100-50200) or any free port (if the value is 0), '[jobmanager|taskmanager].rpc.bind-port' accepts only one specific non-zero port. 

Minor: I'm actually thinking about changing 'taskmanager.host' to 'taskmanager.rpc.address', with backward compatibility on the old key, for unifying the config keys' format. Not absolutely necessary though.

> Flink does not work over NAT
> ----------------------------
>
>                 Key: FLINK-15911
>                 URL: https://issues.apache.org/jira/browse/FLINK-15911
>             Project: Flink
>          Issue Type: Bug
>          Components: Runtime / Coordination
>    Affects Versions: 1.10.0
>            Reporter: Till Rohrmann
>            Assignee: Xintong Song
>            Priority: Blocker
>              Labels: usability
>             Fix For: 1.11.0
>
>
> Currently, it is not possible to run Flink over network address translation. The problem is that the Flink processes do not allow to specify separate bind and external ports. Moreover, the {{TaskManager}} tries to resolve the given {{taskmanager.host}} which might not be resolvable. This breaks NAT or docker setups where the external address is not resolvable from within the container/internal network.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)