RFR [14] 8151678: com/sun/jndi/ldap/LdapTimeoutTest.java failed due to timeout on DeadServerNoTimeoutTest is incorrect

[Pavel Rappo]

Updated,

    http://cr.openjdk.java.net/~prappo/8151678/webrev.01/

> On 30 Aug 2019, at 12:28, Daniel Fuchs <daniel.fuchs at oracle.com> wrote:
> 
> Hi Pavel,
> 
> Looks probably OK ;-)
> 
> One thing though is that it's not always guaranteed that
> InetAddress.getByName("localhost") will resolve to the same
> address than that returned by InetAddess.getLoopbackAddress(),
> so it may be better to use the literal address (which may be
> an IPv6 literal requiring enclosing "[" "]") rather than
> the hostname when building the LDAP URL.
> 
> best regards,
> 
> -- daniel
> 
> On 30/08/2019 11:53, Pavel Rappo wrote:
>>                         //
>>                         // Dear maintainer:
>>                         //
>>                         // Once you are done trying to 'optimize' this routine,
>>                         // and have realized what a terrible mistake that was,
>>                         // please increment the following counter as a warning
>>                         // to the next guy:
>>                         //
>>                         // total_hours_wasted_here = 42
>>                         //
>>                                                                         --- [1]
>> Hello,
>> Please review the following change:
>>      http://cr.openjdk.java.net/~prappo/8151678/webrev.00/
>> Testing timeouts is not an easy task. The test in question has always been
>> problematic. There have been many attempts to address the unreliability of that
>> test and during its lifetime the test has been tagged (and subsequently
>> untagged) as intermittent and eventually was put into the problem list.
>> The proposed change overhauls the test and patches the code that the test
>> exercises.
>> The code change addresses the multithreading issue associated with 8160768 [2].
>> This issue manifests itself as different runtime exceptions that are thrown
>> rather quickly from calls that the test expects to block for a substantial
>> amount of time. Since that test catches this issue reliably, no new tests are
>> introduced.
>> The overhauled test is excluded from the problem list. Before doing so I had run
>> the test on all the platforms for some 40,000 (4.0E4) times. I have not
>> encountered any failures.
>> And yet I'm under no illusion that this change fixes all the issues. I strongly
>> suspect that the test will eventually fail again. I’ve attempted to prepare for
>> that by adding extra diagnostic outputs and improving the usefulness of traces.
>> -Pavel
>> -------------------------------------------------------------------------------
>> [1] https://stackoverflow.com/questions/184618/what-is-the-best-comment-in-source-code-you-have-ever-encountered
>> [2] 8160768: Add capability to custom resolve host/domain names within the default JNDI LDAP provider
>