You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@lucene.apache.org by "Per Steffensen (JIRA)" <ji...@apache.org> on 2012/12/03 16:43:59 UTC
[jira] [Updated] (SOLR-4140) Collection API: CloudSolrServer cannot
be used to create collection with referencing an existing collection
[ https://issues.apache.org/jira/browse/SOLR-4140?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Per Steffensen updated SOLR-4140:
---------------------------------
Attachment: SOLR-4140_trunk.patch
SOLR-4140_trunk.patch fits on top of trunk (5.x)
Essence of patch:
* In CloudSolrServer, if it is a "/admin/collections" request, use all live-nodes as potential receivers of the request. Do not consider collection provided in request or default-collection
* In BasicDistributedZkTest.testCollectionsAPI, we already create at least 3 new "awholynewcollection_X" collections. Used to be created using HttpSolrServer. Now create the first using a CloudSolrServer without default-collection, and create the second using a CloudSolrServer with a non-existing default-collection. The rest of the collections are still created using HttpSolrServer
** Refactored a little to be able to use the existing code in AbstractFullDistribZkTestBase for CloudSolrServer instance creation. Dont copy code - reuse!
BasicDistributedZkTest is still gree, but didnt run entire test-suite. Couldnt imagine that the fix destroied any existing test. Committer should of course run entire test-suite before commit (asuming you are running test-suite pre-commit). Let me know if something fails, even though I didnt expect it.
> Collection API: CloudSolrServer cannot be used to create collection with referencing an existing collection
> -----------------------------------------------------------------------------------------------------------
>
> Key: SOLR-4140
> URL: https://issues.apache.org/jira/browse/SOLR-4140
> Project: Solr
> Issue Type: Bug
> Components: multicore, SolrCloud
> Affects Versions: 4.0
> Reporter: Per Steffensen
> Assignee: Per Steffensen
> Priority: Minor
> Labels: collection-api, multicore
> Attachments: SOLR-4140_trunk.patch
>
>
> CloudSolrServer needs an existing collection to be able to forward requests. For some requests it is not logical that you ned to have or reference an existing collection. E.g. if you want to use CloudSolrServer to create a collection - especially:
> * If you have no collections already and want to create your first collection
> * If you like to create your CloudSolrServer once and for all, and want to set default-collection on it to point to your favorite collection, but also use this CloudSolrServer to create this favorite collection
> Yes, HttpSolrServer can be used, but
> * Some of us like to use CloudSolrServer for everything
> * Using HttpSolrServer you need to check youself which nodes are running, in order to send the create-request to one of them
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: dev-help@lucene.apache.org