On 5/13/11 3:20 AM, Mark Thomas wrote: > On 13/05/2011 07:25, ma...@apache.org wrote: >> Phil Steitz <phil.ste...@gmail.com> wrote: >> >>> I saw this >>> Failed tests: >>> >>> testMaxTotalInvariant(org.apache.commons.pool2.impl.TestGenericKeyedObjectPool): >>> Thread failed: 2 >>> java.lang.IllegalStateException: Returned object not currently part >>> of this pool >>> at >>> org.apache.commons.pool2.impl.GenericKeyedObjectPool.returnObject(GenericKeyedObjectPool.java:1158) >>> at >>> org.apache.commons.pool2.impl.TestGenericKeyedObjectPool$TestThread.run(TestGenericKeyedObjectPool.java:1582) >>> at java.lang.Thread.run(Thread.java:680) >>> >>> I also get these regularly when running load tests using [performance]. >>> >>> I added some instrumentation to GKOP to dump out >>> objectDeque.getAllObjects().size() and it is always 0 for the key >>> when this happens. >>> >>> Phil > I now have commons-performance running and can reproduce this. Hopefully > a fix will be forthcoming shortly. I've looked once and can't see > anything so as before, any help with this would be appreciated. > I will look into this today. I will also check in the changes that I made last night to [performance] to implement equals(), hashCode() and toString() for the Waiter. Unfortunately, this did not explain the problem, but it made me realize that we should doc the fact that poolable objects need to be hashable from this point forward. My first thought was ensureMinIdle and / or destroy not correctly maintaining allObjects; but that looks OK. Will keep looking.
Phil > Mark > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > For additional commands, e-mail: dev-h...@commons.apache.org > > --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org
