[ 
https://issues.apache.org/jira/browse/SOLR-4790?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Erick Erickson updated SOLR-4790:
---------------------------------

    Attachment: SOLR-4790.patch

Updated version, all tests pass. I took a quick look at what happens if you 
define two cores with the same name old-style, and errors are thrown in 
addition to the warning issued. So leaving it as-is seems fine.

But, while looking at that code I think it throws a bogus warning, it just 
tests for the dataDir defined in the <core> tag. So defining two cores with 
dataDir=./data would issue the warning even though they were in different 
instance dirs.

Two questions: 
1> is 
String absData = new File(instDir, dataDir).getCanonicalPath();
a reliable way to get the full path? If so, the test in this patch will work in 
old-style. It worked on a single test anyway.

2> One could test for duplicate dirs in new-style as well. Is it worth it? Two 
variants come to mind:
2a> Pass a map a-la coreDescriptorMap and do simliar stuff.
2b> in SolrCoreDiscoverer.discover, just analyze the map upon return and issue 
a warning.

Either way it would just be some bookeeping local to SolrCoreDiscoverer and not 
intrusive. But is either one worth the effort?

[~markrmiller] [~andyfowler] any opinions?


                
> When defining a core with the same name (discovery mode or not), 
> CoreContainer should throw an error
> ----------------------------------------------------------------------------------------------------
>
>                 Key: SOLR-4790
>                 URL: https://issues.apache.org/jira/browse/SOLR-4790
>             Project: Solr
>          Issue Type: Improvement
>    Affects Versions: 4.3, 5.0
>            Reporter: Erick Erickson
>            Assignee: Erick Erickson
>         Attachments: SOLR-4790.patch, SOLR-4790.patch
>
>
> When you define a core with the same name as another core (discovery mode 
> definitely, old-style xml probably), last one wins. Which means it's very 
> hard to track down what caused the problem.
> What's worse, the last-encountered core replaces the first one, leading to 
> cores that change an unexpected index.

--
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: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to