[POOL-89] Inconsistency with TestGenericKeyedObjectPool.testMaxTotalLRU

• Type: Bug
• Status: Closed
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: Nightly Builds

Environment

Windows XP, Sun JDK 1.4.2_09

• Assignee: Sandy McArthur Jr
• Reporter: Holger Hoffstätte
• Created: 30/Oct/06 13:33
• Updated: 12/Sep/10 17:43
• Resolved: 31/Oct/06 04:17

Description

While working on POOL-88 I noticed that sometimes testMaxTotalLRU fails:

junit.framework.AssertionFailedError: expected same:<a0> was not:<a5>
at junit.framework.Assert.fail(Assert.java:47)
at junit.framework.Assert.failNotSame(Assert.java:278)
at junit.framework.Assert.assertSame(Assert.java:242)
at junit.framework.Assert.assertSame(Assert.java:249)
at org.apache.commons.pool.impl.TestGenericKeyedObjectPool.testMaxTotalLRU(TestGenericKeyedObjectPool.java:278)

This seems to be a timing issue - Thread.sleep(10) is (in my experience) way too small as interval to be reliable. I first suspected some obscure breakage with my patch but could still reproduce this when simply running testMaxTotalLRU repeatedly against the virgin version.