RFR: JDK-8323964: runtime/Thread/ThreadCountLimit.java fails intermittently on AIX [v2]

David Holmes dholmes at openjdk.org
Tue Jan 23 02:11:27 UTC 2024


On Mon, 22 Jan 2024 14:18:46 GMT, Matthias Baesken <mbaesken at openjdk.org> wrote:

>> Every 1-2 months we run into into such kind of failures on AIX:
>> 
>> 
>> [2.498s][warning][os,thread] Failed to start thread "Unknown thread" - pthread_create failed (11=EAGAIN) for attributes: stacksize: 2112k, guardsize: 0k, detached.
>> [2.498s][warning][os,thread] Number of threads approx. running in the VM: 4223
>> [2.498s][warning][os,thread] Checking JVM parameter MaxExpectedDataSegmentSize (currently 8388608k) might be helpful
>> [2.498s][warning][os,thread] Failed to start the native thread for java.lang.Thread "Thread-4213"
>> 
>> [error occurred during error reporting (), id 0xe0000000, Internal Error (/workspace/openjdk-jdk-dev-aix_ppc64-dbg/jdk/src/hotspot/os/aix/os_aix.cpp:2971)]
>> #
>> # There is insufficient memory for the Java Runtime Environment to continue.
>> # Native memory allocation (malloc) failed to allocate 56 bytes. Error detail: AllocateHeap
>> [thread 3856 also had an error]
>> # An error report file with more information is saved as:
>> # /tmp/hs_err_pid13959464.log
>> [thread 76331 also had an error]
>> 
>> 
>> Seems the MaxExpectedDataSegmentSize (Maximum expected size of the data segment, AIX specific flag)
>> settings in in some cases too low for the test.
>> We experienced this a few times in scenarios where a lot of threads are created (and even warn about this, see os_aix.cpp).
>> So we should better run with a higher than default value of MaxExpectedDataSegmentSize for this test.
>
> Matthias Baesken has updated the pull request incrementally with one additional commit since the last revision:
> 
>   adjust comment about MaxExpectedDataSegmentSize

LGTM.

Thanks

src/hotspot/os/aix/globals_aix.hpp line 56:

> 54:   /*  placing the java heap. If the MaxExpectedDataSegmentSize setting         */   \
> 55:   /*  is too small, we might run into resource issues creating many native     */   \
> 56:   /*  threads, if it is too large, we reduce our chance of getting a low hea   */   \

typo: hea -> heap

-------------

Marked as reviewed by dholmes (Reviewer).

PR Review: https://git.openjdk.org/jdk/pull/17512#pullrequestreview-1837740365
PR Review Comment: https://git.openjdk.org/jdk/pull/17512#discussion_r1462643756


More information about the hotspot-runtime-dev mailing list