Look at the source for the failing tests. What you will likely see
is a lot of threads being started, a delay waiting for them to
complete / hit conditions and then some assertions. If the threads
run into scheduling delays, they may not complete / hit conditions
in time. The timings are generally set so that on a reasonably
loaded machine with decent clock resolution, tests will succeed. It
is not surprising that many will fail on slow or loaded machines.
Adding more delay would likely "fix" this, but the tests already
take a very long time to run.
Phil
On 11/19/18 3:27 PM, Gary Gregory wrote:
HI all:
When I run 'mvn clean verify' on git master, I get test failures on my
rather busy CPU (~60-90%):
[INFO] Running org.apache.commons.pool2.impl.TestGenericKeyedObjectPool
[ERROR] Tests run: 63, Failures: 2, Errors: 0, Skipped: 0, Time elapsed:
177.53 s <<< FAILURE! - in
org.apache.commons.pool2.impl.TestGenericKeyedObjectPool
[ERROR]
testEviction2(org.apache.commons.pool2.impl.TestGenericKeyedObjectPool)
Time elapsed: 15.04 s <<< FAILURE!
java.lang.AssertionError: Should be less than 900 idle, found 962
at
org.apache.commons.pool2.impl.TestGenericKeyedObjectPool.testEviction2(TestGenericKeyedObjectPool.java:519)
[ERROR]
testMinIdleMaxTotalPerKey(org.apache.commons.pool2.impl.TestGenericKeyedObjectPool)
Time elapsed: 0.705 s <<< FAILURE!
java.lang.AssertionError: Should be 5 idle, found 3
at
org.apache.commons.pool2.impl.TestGenericKeyedObjectPool.testMinIdleMaxTotalPerKey(TestGenericKeyedObjectPool.java:672)
Does anyone see this as well?
Gary
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org
For additional commands, e-mail: dev-h...@commons.apache.org
