Mailing-List: contact commits-help@beam.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@beam.apache.org
Date: Tue, 18 Apr 2017 13:08:41 +0000 (UTC)
From: =?utf-8?Q?Jean-Baptiste_Onofr=C3=A9_=28JIRA=29?= <jira@apache.org>
To: commits@beam.apache.org
Message-ID: <JIRA.13064691.1492520276000.305493.1492520921588@Atlassian.JIRA>
In-Reply-To: <JIRA.13064691.1492520276000@Atlassian.JIRA>
References: <JIRA.13064691.1492520276000@Atlassian.JIRA> <JIRA.13064691.1492520276938@jira-lw-us.apache.org>
Subject: [jira] [Commented] (BEAM-1995) JdbcIO.Write randomly fails
 PreparedStatements due to missing variables
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
archived-at: Tue, 18 Apr 2017 13:08:47 -0000


    [ https://issues.apache.org/jira/browse/BEAM-1995?page=3Dcom.atlassian.=
jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=3D15972=
630#comment-15972630 ]=20

Jean-Baptiste Onofr=C3=A9 commented on BEAM-1995:
--------------------------------------------

Interesting. Let me check the {{WriteFn}} and see if the issue is not speci=
fic to {{JdbcIO}} (it could be "global" to any {{DoFn}}).

> JdbcIO.Write randomly fails PreparedStatements due to missing variables
> -----------------------------------------------------------------------
>
>                 Key: BEAM-1995
>                 URL: https://issues.apache.org/jira/browse/BEAM-1995
>             Project: Beam
>          Issue Type: Bug
>          Components: sdk-java-extensions
>            Reporter: Rik Nauta
>            Assignee: Jean-Baptiste Onofr=C3=A9
>
> I'm still investigating but it seems that the JdbcIO.Write fails randomly=
 when submitting batches of mutations with the following types of error mes=
sages ```No value specified for parameter 1```.
> The strange thing is that it works perfectly with the DirectExecution pip=
eline, only when it's submitted to DataFlow does it fail. Also, if I modify=
 the code to not batch requests but immediately commit them (manually, auto=
commit is still off) then it completes without errors.
> I've checked the input data which is correct so it's either an issue with=
 the JDBC driver or I suspect that there might be some kind of concurrency =
issue that only shows when the pipeline is executed parallel? Maybe the Pre=
paredStatement should be transient or something like in other PTransforms?
> As I said, I'm still investigating but would love some assistance.


--
This message was sent by Atlassian JIRA
(v6.3.15#6346)