You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@activemq.apache.org by "Justin Bertram (JIRA)" <ji...@apache.org> on 2016/10/04 18:14:20 UTC

[jira] [Commented] (ARTEMIS-768) Broker delivers STOMP message with invalid payload when using core api ClientProducer

    [ https://issues.apache.org/jira/browse/ARTEMIS-768?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15546194#comment-15546194 ] 

Justin Bertram commented on ARTEMIS-768:
----------------------------------------

Can you provide a reproducible test-case for this?  I just added a simple test to org.apache.activemq.artemis.tests.integration.stomp.StompTest in an attempt to reproduce this, but everything looks OK to me.

> Broker delivers STOMP message with invalid payload when using core api ClientProducer
> -------------------------------------------------------------------------------------
>
>                 Key: ARTEMIS-768
>                 URL: https://issues.apache.org/jira/browse/ARTEMIS-768
>             Project: ActiveMQ Artemis
>          Issue Type: Bug
>          Components: Stomp
>    Affects Versions: 1.1.0
>         Environment: Redhat Linux 2.6
>            Reporter: Ruben Cala
>
> When using core api ClientProducer to publish messages with a text payload (xml string, btw) to the broker, the messages sent to stomp consumers have an invalid stomp body.
> In the message delivered to stomp consumers, the  original text payload was 5 characters longer than the published payload.  The stomp messages delivered had 3 extra characters at the front of the payload, and 2 extra characters at the end (spaces), before the final terminating NULL character.
> Besides the fact that the string payload is different from what was published, the 3 extra characters in the front of the stomp message often contained the NULL character.
> This violates the stomp MESSAGE format, and was causing incorrect stomp message parsing.
> The workaround for this was to use the stomp content-length header to determine how many bytes to read in for the body, then manually discarding the garbage characters.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)