You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@lucene.apache.org by "Manoj Lawrence (JIRA)" <ji...@apache.org> on 2016/11/11 22:58:58 UTC

[jira] [Comment Edited] (SOLR-9477) UpdateRequestProcessors ignore child documents

    [ https://issues.apache.org/jira/browse/SOLR-9477?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15658432#comment-15658432 ] 

Manoj Lawrence edited comment on SOLR-9477 at 11/11/16 10:57 PM:
-----------------------------------------------------------------

[~arafalov] This is my first involvement in the JIRA board here. If this is not where we add use cases for the above bug, let me know I will move it to the appropriate location.

We use dataimporthandler to fetch data from DB2. The tablespace is limited in our organisation (not under my control), so if I join couple of large tables it can't handle more than 1 million records.
So, we wanted to split the queries based on certain column values. And to avoid repeating the query in mutiple entities, with different where clauses, I decided to use DIH's ability to have nested entities. This way I am able to split the query into some 20 parts for 12 million records.

That's where this issue with URP not supporting child entity fields is proving to be roadblock for my solution.


was (Author: razen):
[~arafalov] This is my first involvement in the JIRA board here. If this not where we add use cases for the above bug, let me know I will move it to the appropriate location.

We use dataimporthandler to fetch data from DB2. The tablespace is limited in our organisation (not under my control), so if I join couple of large tables it can't handle more than 1 million records.
So, we wanted to split the queries based on certain column values. And to avoid repeating the query in mutiple entities, with different where clauses, I decided to use DIH's ability to have nested entities. This way I am able to split the query into some 20 parts for 12 million records.

That's where this issue with URP not supporting child entity fields is proving to be roadblock for my solution.

> UpdateRequestProcessors ignore child documents
> ----------------------------------------------
>
>                 Key: SOLR-9477
>                 URL: https://issues.apache.org/jira/browse/SOLR-9477
>             Project: Solr
>          Issue Type: Bug
>      Security Level: Public(Default Security Level. Issues are Public) 
>    Affects Versions: 6.2, master (7.0)
>            Reporter: Alexandre Rafalovitch
>            Assignee: Alexandre Rafalovitch
>              Labels: UpdateProcessor
>
> UpdateRequestProcessors completely ignore child documents. The only exception is AddSchemaFieldsUpdateProcessorFactory. The rest seem to be completely unaware that SolrInputDocument has getChildDocuments() or related methods.
> Easy test (on Solr 6.2):
> This works (with IDs auto-assigned and field names generated):
> {code}
> bin/solr create -c childtest
> bin/post -c childtest -type application/json -format solr -d '[{"a":1,"b":2}]'
> {code}
> This fails as the second/third command, with "missing ID field":
> {code}
> bin/post -c childtest -type application/json -format solr -d '[{"a":1,"b":2,"_childDocuments_":[{"c":3,"d":4}]}]'
> {code}
> The message:
> {noformat}
> SimplePostTool version 5.0.0
> POSTing args to http://localhost:8983/solr/childtest/update...
> SimplePostTool: WARNING: Solr returned an error #400 (Bad Request) for url: http://localhost:8983/solr/childtest/update
> SimplePostTool: WARNING: Response: {"responseHeader":{"status":400,"QTime":4},"error":{"metadata":["error-class","org.apache.solr.common.SolrException","root-error-class","org.apache.solr.common.SolrException"],"msg":"[doc=null] missing required field: id","code":400}}
> SimplePostTool: WARNING: IOException while reading response: java.io.IOException: Server returned HTTP response code: 400 for URL: http://localhost:8983/solr/childtest/update
> COMMITting Solr index changes to http://localhost:8983/solr/childtest/update...
> Time spent: 0:00:00.042
> {noformat}
> I also verified it with BlankRemoving URP. I think this is a global problem.



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

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: dev-help@lucene.apache.org