奇怪的master_peer启动问题

阅读更多

这几天在suse10下面配置oracle的enterprise manager的时候，master peer起不来，在rhel3下面一点问题都没有
起来的时候出错：

stat:/opt/oracle/product/9.2.0/network/snmp/peer # ./master_peer CONFIG.master NOV
Setup select/poll problem: Bad file descriptor
./master_peer: error processing configuration
*** glibc detected *** ./master_peer: double free or corruption (fasttop): 0x0807d948 ***
======= Backtrace: =========
/lib/libc.so.6[0xb7e40cb1]
/lib/libc.so.6(__libc_free+0x84)[0xb7e42314]
./master_peer[0x805cd09]
./master_peer[0x804e120]
./master_peer[0x804b275]
/lib/libc.so.6(__libc_start_main+0xdc)[0xb7df287c]
./master_peer(_IO_getc+0x41)[0x8048dc1]
======= Memory map: ========
08048000-08066000 r-xp 00000000 68:03 97956      /opt/oracle/product/9.2.0/network/snmp/peer/master_peer
08066000-08067000 rwxp 0001d000 68:03 97956      /opt/oracle/product/9.2.0/network/snmp/peer/master_peer
08067000-0809e000 rwxp 08067000 00:00 0          [heap]
b7b9c000-b7bd1000 r-xs 00000000 68:02 1183412    /var/run/nscd/dbVtklvZ (deleted)
b7bd1000-b7bd9000 r-xp 00000000 68:02 738756     /lib/libnss_files-2.4.so
b7bd9000-b7bdb000 rwxp 00007000 68:02 738756     /lib/libnss_files-2.4.so
b7c00000-b7c21000 rwxp b7c00000 00:00 0
b7c21000-b7d00000 ---p b7c21000 00:00 0
b7dd1000-b7ddb000 r-xp 00000000 68:02 738781     /lib/libgcc_s.so.1
b7ddb000-b7ddc000 rwxp 00009000 68:02 738781     /lib/libgcc_s.so.1
b7ddc000-b7ddd000 rwxp b7ddc000 00:00 0
b7ddd000-b7ef8000 r-xp 00000000 68:02 738739     /lib/libc-2.4.so
b7ef8000-b7efa000 r-xp 0011a000 68:02 738739     /lib/libc-2.4.so
b7efa000-b7efc000 rwxp 0011c000 68:02 738739     /lib/libc-2.4.so
b7efc000-b7eff000 rwxp b7efc000 00:00 0
b7f0b000-b7f0c000 rwxp b7f0b000 00:00 0
b7f0c000-b7f26000 r-xp 00000000 68:02 738732     /lib/ld-2.4.so
b7f26000-b7f28000 rwxp 0001a000 68:02 738732     /lib/ld-2.4.so
bff5e000-bff73000 rwxp bff5e000 00:00 0          [stack]
ffffe000-fffff000 ---p 00000000 00:00 0          [vdso]
Aborted (core dumped)
stat:/opt/oracle/product/9.2.0/network/snmp/peer #

查看core文件

stat:/opt/oracle/product/9.2.0/network/snmp/peer # gdb master_peer core
GNU gdb 6.6
Copyright (C) 2006 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for details.
This GDB was configured as "i586-suse-linux"...
Using host libthread_db library "/lib/libthread_db.so.1".

warning: Can't read pathname for load map: Input/output error.
Reading symbols from /lib/libc.so.6...done.
Loaded symbols for /lib/libc.so.6
Reading symbols from /lib/ld-linux.so.2...done.
Loaded symbols for /lib/ld-linux.so.2
Reading symbols from /lib/libnss_files.so.2...done.
Loaded symbols for /lib/libnss_files.so.2
Reading symbols from /lib/libgcc_s.so.1...done.
Loaded symbols for /lib/libgcc_s.so.1
Core was generated by `./master_peer CONFIG.master NOV'.
Program terminated with signal 6, Aborted.
#0  0xffffe410 in __kernel_vsyscall ()
(gdb) where
#0  0xffffe410 in __kernel_vsyscall ()
#1  0xb7e058d0 in raise () from /lib/libc.so.6
#2  0xb7e06ff3 in abort () from /lib/libc.so.6
#3  0xb7e3b32b in __libc_message () from /lib/libc.so.6
#4  0xb7e40cb1 in malloc_printerr () from /lib/libc.so.6
#5  0xb7e42314 in free () from /lib/libc.so.6
#6  0x0805cd09 in mgmt_term_env ()
#7  0x0804e120 in agent_shutdown ()
#8  0x0804b275 in main ()
(gdb)

进程起来的时候怎么调用了agent_shutdown 函数？

master_peer.out显示错误：

Setup select/poll problem: Bad file descriptor   

./master_peer: error processing configuration

在metalink下找到一个文章，需要设置把ulimit -n 2048，尝试设置成2048，还是不行，实在没办法，推出crt登陆，直接进root下，再尝试启动master_peer，居然起来了

观察了一下，直接进root的时候ulimit -n的值是1024，而进oracle后再su到root的话，ulimit -n的值是65536，原来问题真的出在open files上面，metalink还是对的，只是suse10下面要设置1024才行，rhel3上面设置成8192都没问题

是跟操作系统内核还是什么有关？这个open files参数起什么影响？