JDK-8066835 : TEST_BUG: javax/management/remote/mandatory/connection/RMIConnector_NPETest.java fails
  • Type: Bug
  • Component: core-libs
  • Sub-Component: java.rmi
  • Affected Version: 9
  • Priority: P3
  • Status: Closed
  • Resolution: Fixed
  • OS: generic
  • CPU: generic
  • Submitted: 2014-12-06
  • Updated: 2015-06-04
  • Resolved: 2015-01-01
The Version table provides details related to the release that this issue/RFE will be addressed.

Unresolved : Release in which this issue/RFE will be addressed.
Resolved: Release in which this issue/RFE has been resolved.
Fixed : Release in which this issue/RFE has been fixed. The release containing this fix may be available for download as an Early Access Release or a General Availability Release.

To download the current JDK release, click here.
JDK 9
9 b43Fixed
Related Reports
Relates :  
JDK-8067316 - TEST_BUG: update RMI test library with better test.timeout.factor handling
Description
 javax/management/remote/mandatory/connection/RMIConnector_NPETest.java fails on all platforms:
 
    JavaTest Message: Test threw exception: TestFailedException: TEST FAILED: Failed to start rmid, giving up after 159ms.

The reason is that it calls RMID.start() using a timeout period of Long.MAX_VALUE, which causes the deadline timestamp to overflow and wrap around to negative. As a result, the first check against the timestamp seems to indicate a timeout and so the test fails.
Comments
I'm re-closing this bug, for two reasons: 1) in general, one cannot reopen a bug for which a changeset has pushed, since the bug ID cannot be reused for another changeset. 2) This bug has a different root cause from the more recent failures observed in the Hotspot nightly. Explanation follows. This bug (JDK-8066835) is caused by a botched timeout computation in the RMI test library. Note the message says that the timeout occurs after 159ms, which is pretty short. The timeout calculation has been fixed, and that's what this bug covers. Failures with similar error messages have occurred in the Hotspot nightlies. This has previously been observed in JDK-8068161. Note that the message is something like, TEST FAILED: Failed to start rmid, giving up after 17232ms. which is a considerably longer timeout. The default rmid start timeout is 15 seconds, so this is an actual timeout, not an erroneous timeout calculation. Clearly, 15 seconds is still too short if timeouts are occurring repeatedly. That issue in turn is caused by the default 15-second rmid timeout not being scaled in response to the test.timeout.factor factor property, which is set to 8.0 (!) in the Hotspot nightly. Thus, the timeout really should be 120 seconds, not 15 seconds. That's covered by JDK-8067316. A fix for that bug is currently out for review but has stalled because few people are around for the holidays. Since the failures observed in JDK-8068161 have a root cause described by JDK-8067316, I've closed the former as a duplicate of the latter. If it's necessary to track Hotspot test failures, they should be linked to JDK-8067316.
01-01-2015
RULE javax/management/remote/mandatory/connection/RMIConnector_NPETest.java Exception TestFailedException: TEST FAILED: Failed to start rmid, giving up after ...ms.
08-12-2014
After thinking about this a bit I think the RMI test library should be more robust about overflow.
06-12-2014