lucene-solr-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "M. Flatterie" <nicflatte...@yahoo.com>
Subject Re: Migrating from 4.2.1 to 4.3.0
Date Thu, 16 May 2013 18:37:22 GMT
Afternoon, my solr.xml file is the following (and has not changed from 4.2.1 to 4.3.0):

        <Context path="/solr" docBase="/home/tcatadm1/apache-tomcat-7.0.39/webapps/solr.war"
debug="0" crossContext="true">
            <Environment name="solr/home" type="java.lang.String" value="/home/solradm1"
override="true"/>
        </Context>



________________________________
 From: Mark Miller <markrmiller@gmail.com>
To: solr-user@lucene.apache.org 
Sent: Thursday, May 16, 2013 2:28:52 PM
Subject: Re: Migrating from 4.2.1 to 4.3.0
 

Your solr webapp context appears to be "" rather than "solr". There was a JIRA issue in 4.3
that may have affected this, but I only saw it from a distance, so just a guess.

What does it say in solr.xml for the context (an attribute on <cores>)

- Mark

On May 16, 2013, at 2:02 PM, "M. Flatterie" <nicflatterie@yahoo.com> wrote:

> Greetings, I just started with Solr a couple weeks ago, with version 4.2.1.
> 
> I installed the following setup:
> - ZooKeeper: 3 instances ensemble
> - Solr: on Tomcat, 4 instances
>     - WebOrder_Collection: instances 1 and 2, 1 shard, 1 master, 1 replica
> 
>     - other_collectionA: instances 3 and 4, 1 shard, 1 master, 1 replica
> 
>     - other_CollectionB: instances 3 and 4, 1 shard, 1 master, 1 replica
> 
> 
> With version 4.2.1 everything works fine.  But I do have a problem if I query instance
3 for something in the WebOrder_Collection.  I found that this is a bug in 4.2.1,. I must
query instances 1 or 2 to get results from WebOrder_Collection.
> 
> 
> Now that I have upgraded to 4.3.0 I have the following problem.  My replicas will not
recover.  The recovery will retry, and retry, ... forever.
> 
> Details.  If I look at the Zoo, I see that:
>      - node_name
>             10.0.2.15:8180_solr        in solr 4.2.1
>             10.0.2.15:8180_             in solr 4.3.0
>      - base_url
>            http://10.0.2.15:8180/solr      in solr 4.2.1
> 
>            http://10.0.2.15:8180            in solr 4.3.0
> 
> My solr logs show this:
> 
> 8869687 [RecoveryThread] ERROR org.apache.solr.cloud.RecoveryStrategy  – Error while
trying to recover. core=WebOrder_Collection:org.apache.solr.client.solrj.impl.HttpSolrServer$RemoteSolrException:
Server at http://10.0.2.15:8180 returned non ok status:404, message:Not Found
>     at org.apache.solr.client.solrj.impl.HttpSolrServer.request(HttpSolrServer.java:372)
>     at org.apache.solr.client.solrj.impl.HttpSolrServer.request(HttpSolrServer.java:180)
>     at org.apache.solr.cloud.RecoveryStrategy.sendPrepRecoveryCmd(RecoveryStrategy.java:202)
>     at org.apache.solr.cloud.RecoveryStrategy.doRecovery(RecoveryStrategy.java:346)
>     at org.apache.solr.cloud.RecoveryStrategy.run(RecoveryStrategy.java:223)
> 
> 
> I have not been able to find more info than that.  The Solr cloud diagram shows instance1
as active and leader, instance 2 as recovering.  My solrconfig.xml are identical, except
for the LUCENE_42 or LUCENE_43 tag.
> 
> 
> Any idea?  I hope that it is a configuration issue on my part...
> 
> Thank you for any help, Nic.
Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message