如何从核心转储文件中提取JVM堆转储？

Question

我正在尝试将Linux Java进程的核心转储文件转换为堆转储文件，以便使用Eclipse MAT进行分析。根据this博客文章（已适配新版OpenJDK 12），我创建了一个核心转储文件，然后运行jhsdb jmap将其转换为HPROF格式：

>sudo gcore -o dump 24934
[New LWP 24971]
...
[New LWP 17921]
warning: Could not load shared library symbols for /tmp/jffi4106753050390578111.so.
Do you need "set solib-search-path" or "set sysroot"?
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
0x00007f94c7e9e98d in pthread_join (threadid=140276994615040, thread_return=0x7ffc716d47a8) at pthread_join.c:90
90      pthread_join.c: No such file or directory.
warning: target file /proc/24934/cmdline contained unexpected null characters
warning: Memory read failed for corefile section, 1048576 bytes at 0x7f93756a6000.
warning: Memory read failed for corefile section, 1048576 bytes at 0x7f9379bec000.
...
warning: Memory read failed for corefile section, 1048576 bytes at 0x7f94c82dd000.
Saved corefile dump.24934

> ls -sh dump.24934 
22G dump.24934

> /usr/lib/jvm/zulu-12-amd64/bin/jhsdb jmap --exe /usr/lib/jvm/zulu-12-amd64/bin/java --core dump.24934 --binaryheap --dumpfile jmap-dump.24934
Attaching to core dump.24934 from executable /usr/lib/jvm/zulu-12-amd64/bin/java, please wait...
Debugger attached successfully.
Server compiler detected.
JVM version is 12.0.1+12
null

> ls -sh jmap-dump.24934 
3.3M jmap-dump.24934

核心转储文件为22 Gb，而堆转储文件仅为3 Mb，因此很可能 jhsdb jmap 命令无法处理整个核心转储。此外，Eclipse MAT在打开堆转储文件时出现以下消息：The HPROF parser encountered a violation of the HPROF specification that it could not safely handle. This could be due to file truncation or a bug in the JVM.

- Alex R.

Answer 1

Alex,

这有两种可能性。

首先，gcore是gdb的便捷脚本。我看到它提示一些警告消息，说它难以加载solib。gdb可能一开始就生成了一个破损的核心文件。您可以尝试使用gdb加载核心文件，看看它是否能解析。

其次，jhsdb会自行解析核心文件。您可以使用环境变量LIBSAPROC_DEBUG=1来获取其跟踪信息。它将帮助您了解解析中的问题。

为什么不直接使用jmap -dump转储Java堆？这将跳过核心转储文件。

- Xin Liu