java - JVM 崩溃时 hs_err_pidXXX.log 文件中的 rlimit 信息错误

Question

我有一个使用 jdk 1.8.0_102 运行的 java 应用程序。当应用程序崩溃(内存不足 - 见下文)时，会生成一个文件 hs_err_pidXXX，其中包含有关 jvm、系统、内存、线程等的一些信息...

    #
    # There is insufficient memory for the Java Runtime Environment to continue.
    # Native memory allocation (mmap) failed to map 65536 bytes for committing reserved memory.
    # Possible reasons:
    #   The system is out of physical RAM or swap space
    #   In 32 bit mode, the process size limit was hit
    # Possible solutions:
    #   Reduce memory load on the system
    #   Increase physical memory or swap space
    #   Check if swap backing store is full
    #   Use 64 bit Java on a 64 bit OS
    #   Decrease Java heap size (-Xmx/-Xms)
    #   Decrease number of Java threads
    #   Decrease Java thread stack sizes (-Xss)
    #   Set larger code cache with -XX:ReservedCodeCacheSize=
    # This output file may be truncated or incomplete.
    #
    #  Out of Memory Error (os_linux.cpp:2627), pid=1094, tid=0x00007fac4041e700
    #
    # JRE version: Java(TM) SE Runtime Environment (8.0_102-b14) (build 1.8.0_102-b14)
    # Java VM: Java HotSpot(TM) 64-Bit Server VM (25.102-b14 mixed mode linux-amd64 compressed oops)
    # Failed to write core dump. Core dumps have been disabled. To enable core dumping, try "ulimit -c unlimited" before starting Java again
    #

...

    ---------------  S Y S T E M  ---------------

    OS:DISTRIB_ID=Ubuntu
    DISTRIB_RELEASE=14.04
    DISTRIB_CODENAME=trusty
    DISTRIB_DESCRIPTION="Ubuntu 14.04.5 LTS"

    uname:Linux 3.13.0-101-generic #148-Ubuntu SMP Thu Oct 20 22:08:32 UTC 2016 x86_64
    libc:glibc 2.19 NPTL 2.19 
    rlimit: STACK 8192k, CORE 0k, NPROC 30034, NOFILE 4096, AS infinity
    load average:10.38 6.50 2.88

In the list of information above, i can see the following line: rlimit: STACK 8192k, CORE 0k, NPROC 30034, NOFILE 4096, AS infinity

My question is the following: why NOFILE does not correspond to the value I have set on my system (in limits.conf file, it should be 20000)? When I run the command ulimit -n with the same user as the one running the jvm, i have a different value. Note that STACK displayed is the correct value that I have set on my system, not the default one. JVM is run on AWS c3.large on demand instance.

Here is the result of the ulimit -a command :

core file size          (blocks, -c) 0
data seg size           (kbytes, -d) unlimited
scheduling priority             (-e) 0
file size               (blocks, -f) unlimited
pending signals                 (-i) 30034
max locked memory       (kbytes, -l) 64
max memory size         (kbytes, -m) unlimited
open files                      (-n) 20000
pipe size            (512 bytes, -p) 8
POSIX message queues     (bytes, -q) 819200
real-time priority              (-r) 0
stack size              (kbytes, -s) 8192
cpu time               (seconds, -t) unlimited
max user processes              (-u) 30034
virtual memory          (kbytes, -v) unlimited
file locks                      (-x) unlimited

Answer 1

我检查了源代码¹。事实证明，转储中打印的 rlimit 值是通过调用 getrlimit 获取的。这是一个 POSIX 库方法，直接从操作系统获取信息。 JVM 不会减少其 NOFILES rlimit，但在某些情况下，它会将软 rlimit 增加到硬 rlimit 的值。

因此，我怀疑转储中报告的 rlimit必须在启动运行应用程序的 JVM 的环境中设置。这不能与运行 ulimit 的 shell 环境相同。

<小时/>

^{1 - ...使用 find 和 grep，所以我可能错过了一些东西。}