You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@lucene.apache.org by "Hoss Man (JIRA)" <ji...@apache.org> on 2015/04/25 01:59:38 UTC
[jira] [Updated] (SOLR-7470) jvm/filesystem dependent failures in
SolrCloudExampleTest.testLoadDocsIntoGettingStartedCollection -- directory
traversal order dependency
[ https://issues.apache.org/jira/browse/SOLR-7470?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Hoss Man updated SOLR-7470:
---------------------------
Attachment: SOLR-7470.patch
Attached patch:
* fixes the test to index the xml files in a deterministic order dictated by the test seed.
* fixes mem.xml to use float price value.
If you apply the patch, and then revert the mem.xml change, anyone -- regardless of filesystem -- should be able to see the test reliably fail with this seed...
{noformat}
ant test -Dtestcase=SolrCloudExampleTest -Dtests.method=testLoadDocsIntoGettingStartedCollection -Dtests.seed=2AD197B874223638:C770325C8CC3DD07 -Dtests.slow=true -Dtests.locale=es_PE -Dtests.timezone=Africa/Khartoum -Dtests.asserts=true -Dtests.file.encoding=UTF-8
{noformat}
...i'm still hammering on the test to verify there are no other file orderings that trigger similar bugs.
> jvm/filesystem dependent failures in SolrCloudExampleTest.testLoadDocsIntoGettingStartedCollection -- directory traversal order dependency
> ------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: SOLR-7470
> URL: https://issues.apache.org/jira/browse/SOLR-7470
> Project: Solr
> Issue Type: Bug
> Reporter: Hoss Man
> Assignee: Hoss Man
> Fix For: 5.2
>
> Attachments: SOLR-7470.patch
>
>
> SolrCloudExampleTest.testLoadDocsIntoGettingStartedCollection currently fails 100% reliably on my laptop, regardless of seed with a root cause of...
> {noformat}
> [junit4] 2> 35968 T60 N:127.0.0.1:57372_ C:gettingstarted S:shard1 R:core_node1 c:gettingstarted_shard1_replica2 C15 oasc.SolrException.log ERROR org.apache.solr.common.SolrException: ERROR: [doc=VS1GB400C3] Error adding field 'price'='74.99' msg=For input string: "74.99"
> [junit4] 2> at org.apache.solr.update.DocumentBuilder.toDocument(DocumentBuilder.java:178)
> [junit4] 2> at org.apache.solr.update.AddUpdateCommand.getLuceneDocument(AddUpdateCommand.java:79)
> [junit4] 2> at org.apache.solr.update.DirectUpdateHandler2.addDoc0(DirectUpdateHandler2.java:237)
> [junit4] 2> at org.apache.solr.update.DirectUpdateHandler2.addDoc(DirectUpdateHandler2.java:163)
> [junit4] 2> at org.apache.solr.update.processor.RunUpdateProcessor.processAdd(RunUpdateProcessorFactory.java:69)
> {noformat}
> This test uses "data_driven_schema_configs" and indexes all of the \*.xml files in "example/exampledocs".
> Two factors seem to be contributing to the the reason this fails consistently for me (but not for jenkins or anyone else i've seen complain)...
> # The {{File.listFiles(FileFilter)}} method is used to iterate over the files
> # The "mem.xml" file has an integer price value: {{<field name="price">185</field>}}
> {{listFiles}} is documented that "There is no guarantee that the name strings in the resulting array will appear in any specific order" and evidently with my filesystem + JVM they come back in a consistent order everytime, which just so happens to put mem.xml in front of any other file that also has a "price" field.
--
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