8221303: sun/management/jmxremote/bootstrap/JMXInterfaceBindingTest.java fails due to java.rmi.server.ExportException: Port already in use

Alex Menkov alexey.menkov at oracle.com
Thu Jul 18 17:44:05 UTC 2019


+1

--alex

On 07/18/2019 08:27, Chris Plummer wrote:
> Hi Daniil,
> 
> Looks good.
> 
> Chris
> 
> On 7/17/19 6:26 PM, Daniil Titov wrote:
>> Hi Chris,
>>
>> Yes, I added output.reportDiagnosticSummary() in webrev-01, but 
>> removed it
>> In webrev-02, and later restored it in webrev-03.
>>
>> When the test runs of retries output.getExitValue() is set to 1 
>> (COMMUNICATION_ERROR_EXIT_VAL)
>>   (the "exitValue =1 " in the previous email).
>>
>> Please review a new version of the fix that has an explicit error 
>> message printed
>> if the test runs out of retries (line 168).
>>
>> Webrev: http://cr.openjdk.java.net/~dtitov/8221303/webrev.04/
>> Bug: https://bugs.openjdk.java.net/browse/JDK-8221303
>>
>> Thanks!
>> --Daniil
>>
>> On 7/17/19, 5:33 PM, "Chris Plummer" <chris.plummer at oracle.com> wrote:
>>
>>      Hi Daniil,
>>      I'm confused now. I mentioned output.reportDiagnosticSummary() 
>> because I
>>      thought I saw it in your webrev-02. Now I don't. Maybe I had glanced
>>      back at webrev-01 and saw it there. One issue with exceptions in the
>>      output that are not considered errors is that if later there is an
>>      error, mdash wills show the exception as one of (possible multiple)
>>      reasons for the failure, so it's good to avoid if possible. Looks 
>> like
>>      what you have now is ok.
>>      I have a question about what happens if you run out of retries. 
>> What is
>>      output.getExitValue() set to? Also, I think you should print an an
>>      explicit error message indicating you've run out of retries.
>>      thanks,
>>      Chris
>>      On 7/17/19 4:36 PM, Daniil Titov wrote:
>>      > Hi Chris,
>>      >
>>      > output.reportDiagnosticSummary() prints the output from the 
>> process (both stdout and stderr) and the exit value to the test's stderr.
>>      > In case if the port is already in use it prints the following:
>>      >
>>      > stdout: [];
>>      >   stderr: [Error: JMX connector server communication error: 
>> service:jmx:rmi://127.0.0.1:9101
>>      > jdk.internal.agent.AgentConfigurationError: 
>> java.rmi.server.ExportException: Port already in use: 9101; nested 
>> exception is:
>>      >     java.net.BindException: Address already in use
>>      >     at 
>> jdk.management.agent/sun.management.jmxremote.ConnectorBootstrap.exportMBeanServer(ConnectorBootstrap.java:820) 
>>
>>      >     at 
>> jdk.management.agent/sun.management.jmxremote.ConnectorBootstrap.startRemoteConnectorServer(ConnectorBootstrap.java:479) 
>>
>>      >     at 
>> jdk.management.agent/jdk.internal.agent.Agent.startAgent(Agent.java:447)
>>      >     at 
>> jdk.management.agent/jdk.internal.agent.Agent.startAgent(Agent.java:599)
>>      > Caused by: java.rmi.server.ExportException: Port already in 
>> use: 9101; nested exception is:
>>      > < I skipped the rest>
>>      > ]
>>      >   exitValue = 1
>>      >
>>      > It makes sense to have it called if the test fails, otherwise 
>> this information would be missed in the test output.
>>      > Please review the new version of the fix that has the call to 
>> output.reportDiagnosticSummary() restored.
>>      >
>>      > Thanks!
>>      >
>>      > Webrev: http://cr.openjdk.java.net/~dtitov/8221303/webrev.03/
>>      > Bug: https://bugs.openjdk.java.net/browse/JDK-8221303
>>      >
>>      > -Daniil
>>      >
>>      > On 7/17/19, 3:58 PM, "Chris Plummer" 
>> <chris.plummer at oracle.com> wrote:
>>      >
>>      >      What does output.reportDiagnosticSummary() print out then 
>> the port is
>>      >      already in use, and have you made this happen with your 
>> fixes in place?
>>      >
>>      >      Chris
>>      >
>>      >      On 7/17/19 3:20 PM, Daniil Titov wrote:
>>      >      > Hi Chris and Alex,
>>      >      >
>>      >      > Please review a new version of the fix that moves the 
>> diagnostic output for the test failure to run()
>>      >      > method after the number of retry attempts is exceeded. 
>> It also includes other corrections that
>>      >      > you suggested.
>>      >      >
>>      >      > Thanks!
>>      >      >
>>      >      > Webrev: 
>> http://cr.openjdk.java.net/~dtitov/8221303/webrev.02/
>>      >      > Bug: https://bugs.openjdk.java.net/browse/JDK-8221303
>>      >      >
>>      >      > --Best regards,
>>      >      > Daniil
>>      >      >
>>      >      >
>>      >      > On 7/17/19, 1:47 PM, "Chris Plummer" 
>> <chris.plummer at oracle.com> wrote:
>>      >      >
>>      >      >      Hi Daniil,
>>      >      >
>>      >      >      I think you can remove "Ok" from the following 
>> message:
>>      >      >
>>      >      >        237 System.out.println("DEBUG: OK. Spawned java 
>> process terminated
>>      >      >      with expected exit code of "
>>      >      >        238                         + 
>> STOP_PROCESS_EXIT_VAL);
>>      >      >
>>      >      >      It's somewhat misleading since the test can still 
>> fail. I think the
>>      >      >      following is also misleading:
>>      >      >
>>      >      >        249             if (testFailed) {
>>      >      >        250                 
>> output.reportDiagnosticSummary();
>>      >      >        251                 if 
>> (output.getStderr().contains("Port already in
>>      >      >      use")) {
>>      >      >        252                     // Need to retry
>>      >      >        253                     return true;
>>      >      >        254                 }
>>      >      >        255             }
>>      >      >
>>      >      >      The test can still pass after this happens, right? 
>> And I think there are
>>      >      >      other "Test FAILURE" error message that can appear 
>> just before this.
>>      >      >      Perhaps the code above should add a println that 
>> indicates that there
>>      >      >      will be a retry attempt since the failure was due 
>> to the port being in use.
>>      >      >
>>      >      >      Otherwise looks good.
>>      >      >
>>      >      >      thanks,
>>      >      >
>>      >      >      Chris
>>      >      >
>>      >      >      On 7/17/19 1:30 PM, Alex Menkov wrote:
>>      >      >      > Hi Daniil,
>>      >      >      >
>>      >      >      > The fix looks good in general.
>>      >      >      > Couple cosmetic notes (no new webrev required):
>>      >      >      >
>>      >      >      >  162                 needRetry = runTest();
>>      >      >      >  163             }
>>      >      >      >  164             while (needRetry && (attempts++ 
>> < MAX_RETRY_ATTEMTS));
>>      >      >      > Please move "while" to the prev line:
>>      >      >      >  163             } while (needRetry && 
>> (attempts++ < MAX_RETRY_ATTEMTS));
>>      >      >      >
>>      >      >      >
>>      >      >      >  242                     System.out.println("Test 
>> FAILURE on" + name +
>>      >      >      > " reason: The expected line \"" + READY_MSG
>>      >      >      >  243                             + "\" is not 
>> present in the process
>>      >      >      > output");
>>      >      >      > Please add space: "Test FAILURE on " + name
>>      >      >      >
>>      >      >      > --alex
>>      >      >      >
>>      >      >      > On 07/17/2019 12:46, Daniil Titov wrote:
>>      >      >      >> Hi Chris,
>>      >      >      >>
>>      >      >      >>>     It's a little unclear to me why you moved 
>> from ProcessThread to
>>      >      >      >>>    TestProcessThread + Process. An explanation 
>> of that would make it
>>      >      >      >>> easier
>>      >      >      >>>    to understand many of the changes.
>>      >      >      >>
>>      >      >      >> There are two reasons for that:
>>      >      >      >> 1)  For every network interface the test starts 
>> a separate thread
>>      >      >      >> that runs the test for this interface. We want 
>> that if the test fails
>>      >      >      >>      due to the bind error the thread not to 
>> exit but try to repeat
>>      >      >      >> the test several times. 
>> jdk.test.lib.thread.ProcessThread doesn't
>>      >      >      >> allow this.
>>      >      >      >> 2) To filter out the cases when the test fails 
>> due to the bind error
>>      >      >      >> we need to parse the process output. 
>> jdk.test.lib.thread.ProcessThread
>>      >      >      >>    registers its own pumps for stdin and sdtout  
>> (by calling
>>      >      >      >> ProcessTools.startProcess()) that consume all 
>> output of the process
>>      >      >      >> and prevent
>>      >      >      >>    the output analyzer from collecting this output.
>>      >      >      >>   Thanks!
>>      >      >      >> --Daniil
>>      >      >      >>
>>      >      >      >> On 7/17/19, 12:11 PM, "Chris Plummer" 
>> <chris.plummer at oracle.com> wrote:
>>      >      >      >>
>>      >      >      >>      Hi Daniil,
>>      >      >      >>           It's a little unclear to me why you 
>> moved from
>>      >      >      >> ProcessThread to
>>      >      >      >>      TestProcessThread + Process. An explanation 
>> of that would make
>>      >      >      >> it easier
>>      >      >      >>      to understand many of the changes.
>>      >      >      >>           thanks,
>>      >      >      >>           Chris
>>      >      >      >>           On 7/11/19 10:16 AM, Daniil Titov wrote:
>>      >      >      >>      > Please review the change that fixes an 
>> intermittent failure of
>>      >      >      >> 
>> sun/management/jmxremote/bootstrap/JMXInterfaceBindingTest.java
>>      >      >      >>      > test due to ports collision. The tests 
>> finds all network
>>      >      >      >> interfaces and for every interface starts a 
>> separate process that tests
>>      >      >      >>      > the connection to JMX agent server for a 
>> specific
>>      >      >      >> ports/interface combination.
>>      >      >      >>      >
>>      >      >      >>      > The test was changed to retry in case of 
>> the failure. If the
>>      >      >      >> subprocess fails to bind and the number
>>      >      >      >>      > of retry attempts doesn't exceed the 
>> limit a new pair of
>>      >      >      >> random ports is selected and the test is run again.
>>      >      >      >>      >
>>      >      >      >>      > Webrev: 
>> http://cr.openjdk.java.net/~dtitov/8221303/webrev.01/
>>      >      >      >>      > Bug: 
>> https://bugs.openjdk.java.net/browse/JDK-8221303
>>      >      >      >>      >
>>      >      >      >>      > Thanks!
>>      >      >      >>      > --Daniil
>>      >      >      >>      >
>>      >      >      >>      >
>>      >      >      >>
>>      >      >      >>
>>      >      >
>>      >      >
>>      >      >
>>      >      >
>>      >
>>      >
>>      >
>>      >
>>      >
>>
>>
> 
> 


More information about the serviceability-dev mailing list