Mailing-List: contact users-help@archiva.apache.org; run by ezmlm
Precedence: bulk
Reply-To: users@archiva.apache.org
Received-SPF: pass (athena.apache.org: domain of odeaching@gmail.com
 designates 209.85.221.178 as permitted sender)
DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=gmail.com; s=gamma;
        h=mime-version:sender:in-reply-to:references:date
         :x-google-sender-auth:message-id:subject:from:to:content-type;
        b=Q9waRt6WSWIPgcuK9zE1Vd6/h2Dn0b9MZ7HEbYYDS5Z/PaXWpRloUwi8Tpbw/1mOYO
         UBxrOGDsiMKCSce/724UAzrMJSiro/tmfS9ZscqWRk8Lc0Odh52fM3049Q30sDyej7B4
         nFIgK+IpwCqYlRDEZ91fri5tstDwjAk5m571w=
MIME-Version: 1.0
Sender: odeaching@gmail.com
In-Reply-To: <23865774.post@talk.nabble.com>
References: <23865774.post@talk.nabble.com>
Date: Fri, 5 Jun 2009 19:09:50 +0800
Message-ID: <8667b1bd0906050409q68289737gb8b593c1c01fa893@mail.gmail.com>
Subject: Re: upgrade to 1.2.1 => numerous SQL errors
From: Deng Ching <oching@apache.org>
To: users@archiva.apache.org
Content-Type: multipart/alternative; boundary=00163646d979693a8a046b97ee00

--00163646d979693a8a046b97ee00
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

Hi Marc,

On Thu, Jun 4, 2009 at 4:25 PM, Marc Lustig <ml@marclustig.com> wrote:

>
> On startup, Archiva 1.2.1 runs the repo-scanner and throws endless
> exceptions.
>
> The first type of exception looks like this:
>
> 2009-06-04 03:03:11,727 [pool-2-thread-1] ERROR
>
> org.apache.maven.archiva.repository.scanner.functors.ConsumerProcessFileClosure
> - Consumer [repository-purge] had an error when processing file
> [/opt/managed_repos/snapshots/.in
> dex/nexus-maven-repository-index.zip]: Not enough parts to the path
> [.index/nexus-maven-repository-index.zip] to construct an ArchivaArtifact
> from. (Requires at least 4 parts)
> org.apache.maven.archiva.consumers.ConsumerException: Not enough parts to
> the path [.index/nexus-maven-repository-index.zip] to construct an
> ArchivaArtifact from. (Requires at least 4 parts)
>

This is a known issue, see http://jira.codehaus.org/browse/MRM-1151 :)


>
>
> the second type is sql-related:
>
> 2009-06-04 09:07:53,420 [pool-2-thread-1] ERROR
>
> org.apache.maven.archiva.repository.scanner.functors.ConsumerProcessFileClosure
> - Consumer [update-db-artifact] had an error when processing file
> [/opt/managed_repos/internal/de
> /company/azcommons/azc-services/1.0.35/azc-services-1.0.35.pom]: Insert
> request failed: INSERT INTO ARCHIVA.ARCHIVA_ARTIFACT
>
> (SNAPSHOT_VERSION,PLATFORM,ORIGIN,WHEN_PROCESSED,CHECKSUM_SHA1,WHEN_GATHERED,CHECKSUM_MD5,WHEN_INDEXE
>
> D,LAST_MODIFIED,FILE_SIZE,ARTIFACT_ID,CLASSIFIER,GROUP_ID,REPOSITORY_ID,FILE_TYPE,VERSION)
> VALUES (?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?)
> javax.jdo.JDODataStoreException: Insert request failed: INSERT INTO
> ARCHIVA.ARCHIVA_ARTIFACT
>
> (SNAPSHOT_VERSION,PLATFORM,ORIGIN,WHEN_PROCESSED,CHECKSUM_SHA1,WHEN_GATHERED,CHECKSUM_MD5,WHEN_INDEXED,LAST_MODIFIED,FILE_SIZE,ARTIFA
> CT_ID,CLASSIFIER,GROUP_ID,REPOSITORY_ID,FILE_TYPE,VERSION) VALUES
> (?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?)
> ....
> NestedThrowablesStackTrace:
> java.sql.SQLIntegrityConstraintViolationException: ORA-00001: unique
> constraint (ARCHIVA.ARCHIVA_ARTIFACT_PK) violated
>
>
> We have been running Archiva 1.1.3 on Tomcat and Oracle RAC before without
> such errors.
> Before I deployed 1.2.1 I adjusted the JDO-schema (package.jpox), basically
> by adding jdbc-type="clob" to all varchars larger than 250. That strategy
> worked well for us before.
>
>
> Anybody having an idea how to work around those 2 errors?
>
> If not, can I safely delete the database, and let it recreated by Archiva
> on
> startup ?
> (I asked that question already before, but the answer was not clear enough
> for me.)
>

I haven't encountered those errors when I tried the upgrade, but I was
using  the embedded derby database then. To answer your question, yes you
can safely delete the database. I assume you already did that based on your
other email regarding the scanning problem :)

Thanks,
Deng

--00163646d979693a8a046b97ee00--