Mailing-List: contact commits-help@cassandra.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@cassandra.apache.org
Date: Wed, 20 Jul 2016 12:03:20 +0000 (UTC)
From: "Aleksey Yeschenko (JIRA)" <jira@apache.org>
To: commits@cassandra.apache.org
Message-ID: <JIRA.12990807.1468942623000.82234.1469016200649@Atlassian.JIRA>
In-Reply-To: <JIRA.12990807.1468942623000@Atlassian.JIRA>
References: <JIRA.12990807.1468942623000@Atlassian.JIRA> <JIRA.12990807.1468942623535@arcas>
Subject: [jira] [Commented] (CASSANDRA-12236) RTE from new CDC column breaks
 in flight queries.
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
archived-at: Wed, 20 Jul 2016 12:03:22 -0000


    [ https://issues.apache.org/jira/browse/CASSANDRA-12236?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15385732#comment-15385732 ] 

Aleksey Yeschenko commented on CASSANDRA-12236:
-----------------------------------------------

We went the way we went under the assumption that users would not see this exception unless the made explicit schema changes during the mixed-version period. Looks like you hit the issue even if the user is *not* explicitly performing any schema changes on the cluster, which is arguably less acceptable. My guess is that this is caused by setting up auth, tracing, and system_distributed keyspaces on startup.

Seems to me that the only reasonable way it so add the cdc param conditionally, and have an ugly null default instead of the proper empty frozen set, but, what can you do?

I'll see if that is indeed sufficient and can be done super fast. That said, if it's not, I feel like the fix can wait until 3.9, as I don't believe that this problem is critical enough do delay the already very late 3.8. You are only going to have the session broken once per tuple of (source, destination) nodes while they are upgrading. While not harmless, I don't believe it to be a big enough deal to block an even 3.8 release. 

> RTE from new CDC column breaks in flight queries.
> -------------------------------------------------
>
>                 Key: CASSANDRA-12236
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-12236
>             Project: Cassandra
>          Issue Type: Bug
>            Reporter: Jeremiah Jordan
>            Priority: Blocker
>             Fix For: 3.8
>
>
> This RTE is not harmless. It will cause the internode connection to break which will cause all in flight requests between these nodes to die/timeout.
> {noformat}
>     - Due to changes in schema migration handling and the storage format after 3.0, you will
>       see error messages such as:
>          "java.lang.RuntimeException: Unknown column cdc during deserialization"
>       in your system logs on a mixed-version cluster during upgrades. This error message
>       is harmless and due to the 3.8 nodes having cdc added to their schema tables while
>       the <3.8 nodes do not. This message should cease once all nodes are upgraded to 3.8.
>       As always, refrain from schema changes during cluster upgrades.
> {noformat}


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