[ 
https://issues.apache.org/jira/browse/LUCENE-9331?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17087795#comment-17087795
 ] 

Adrien Grand commented on LUCENE-9331:
--------------------------------------

The following change fixed the issue for me but I'd like to make sure that it 
doesn't defeat the purpose of the test. It increments the amount of free space 
relatively to the current number in order to speed up convergence of the test.

{noformat}
diff --git 
a/lucene/core/src/test/org/apache/lucene/index/TestIndexWriterDelete.java 
b/lucene/core/src/test/org/apache/lucene/index/TestIndexWriterDelete.java
index e879019897b..2e08d3b7ee1 100644
--- a/lucene/core/src/test/org/apache/lucene/index/TestIndexWriterDelete.java
+++ b/lucene/core/src/test/org/apache/lucene/index/TestIndexWriterDelete.java
@@ -702,8 +702,8 @@ public class TestIndexWriterDelete extends LuceneTestCase {
       }
       dir.close();
 
-      // Try again with 10 more bytes of free space:
-      diskFree += 10;
+      // Try again with more bytes of free space:
+      diskFree += Math.max(10, diskFree >>> 3);
     }
     startDir.close();
   }
{noformat}

[~mikemccand] It looks like you are the most familiar with this test, what do 
you think?

> TestIndexWriterDelete.testDeletesOnDiskFull can run for minutes
> ---------------------------------------------------------------
>
>                 Key: LUCENE-9331
>                 URL: https://issues.apache.org/jira/browse/LUCENE-9331
>             Project: Lucene - Core
>          Issue Type: New Feature
>            Reporter: Adrien Grand
>            Priority: Minor
>
> This seed reproduces for me:
> ant test  -Dtestcase=TestIndexWriterDelete 
> -Dtests.method=testUpdatesOnDiskFull -Dtests.seed=63F37B47DD88EF8A 
> -Dtests.nightly=true -Dtests.slow=true -Dtests.badapples=true 
> -Dtests.locale=twq -Dtests.timezone=Greenwich -Dtests.asserts=true 
> -Dtests.file.encoding=ISO-8859-1
> I didn't wait for the test to finish (it looks like it would take a very long 
> time) though logging suggests that it would converge... eventually. This 
> seems to be a combination of slow convergence, mixed with slow features of 
> MockDirectoryWrapper like sleeps when opening inputs and costly calls like 
> checkIndex. 



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org
For additional commands, e-mail: issues-h...@lucene.apache.org

Reply via email to