[jbosscache-dev] Releasing 1.4.1.GA

[Brian Stansberry]

Manik Surtani wrote:
> On 5 Jan 2007, at 16:02, Manik Surtani wrote:
> 
>> On 5 Jan 2007, at 15:37, Brian Stansberry wrote:
>> 
>>> At this point it's completely unclear what their issue is.  I
>>> wouldn't hold up a release for it beyond one more day.
>>> 
>>> Manik, perhaps you can have a look at
>>> 
> o.j.c.statetransfer.VersionedTestBase.testEvictionAfterStateTransfer(
>>> ) and see if you think it adequately tests eviction following state
>>> transfer.  It actually goes beyond that quite a bit now, and tests
>>> eviction in the presence of concurrent creation of new nodes.
>> 
>> This is the only test failing in o.j.c.statetransfer on HEAD at the
>> moment. When was this test added?
> 
> Pls ignore, this is irrelevant.  It passes fine on 1.4.x.
> 
> The test does seem to, as you suggest, test eviction of nodes created
> via state transfer. 
> 

The state transfer tests have been failing on jboss-cache-testsuite-140
since Tues.  There are 2 eviction tests and they weren't cleaning up
after themselves properly, so I believe live caches left over from the
first were impacting the second.  This has been fixed in 1.4.x and will
be in HEAD if it's broken there.