所谓操作系统,是应用程序与服务器硬件进行沟通的中间层。应用程序的所有操作,都是和操作系统进行沟通交互。操作系统负责将所有交互转化为设备语言,进行硬件交互。
我们在进行Oracle故障调试和内核原理工作的时候,经常需要了解后台运行的动作和细节。一些故障场景,如程序进程hang住、无法登陆等问题,就需要操作系统级别监控,检查定位问题。
Oracle自身已经提供了很多这类型的工具,如oradebug、各种等待事件和跟踪方式。此外,各类型的操作系统提供出很多系统级别工具,帮助我们进行监控。本篇主要介绍Linux环境中的strace工具典型使用方法,留待需要朋友待查。
1、环境介绍
笔者使用红帽Red Hat 6.5进行测试。
[root@XXX-Standby-Asm ~]# cat /etc/RedHat-release
Red Hat Enterprise Linux Server release 6.5 (Santiago)
[root@XXX-Standby-Asm ~]# uname -a
Linux XXX-Standby-Asm 2.6.32-431.el6.x86_64 #1 SMP Sun Nov 10 22:19:54 EST 2013 x86_64 x86_64 x86_64 GNU/Linux
对应使用Oracle版本为11.2.0.4。
[oracle@XXX-Standby-Asm ~]$ sqlplus /nolog
SQL*Plus: Release 11.2.0.4.0 Production on Tue Jun 16 16:39:21 2015
Copyright (c) 1982, 2013, Oracle. All rights reserved.
SQL> conn / as sysdba
Connected.
SQL> select * from v$version;
BANNER
--------------------------------------------------------------------------------
Oracle Database 11g Enterprise Edition Release 11.2.0.4.0 - 64bit Production
PL/SQL Release 11.2.0.4.0 - Production
CORE 11.2.0.4.0 Production
TNS for Linux: Version 11.2.0.4.0 - Production
NLSRTL Version 11.2.0.4.0 – Production
2、Strace命令信息
使用-h参数,可以调用出strace命令的语法格式和各个使用参数。
[oracle@XXX-Standby-Asm ~]$ strace -h
usage: strace [-dDffhiqrtttTvVxx] [-a column] [-e expr] ... [-o file]
[-p pid] ... [-s strsize] [-u username] [-E var=val] ...
[command [arg ...]]
or: strace -c [-D] [-e expr] ... [-O overhead] [-S sortby] [-E var=val] ...
[command [arg ...]]
-c -- count time, calls, and errors for each syscall and report summary
-f -- follow forks, -ff -- with output into separate files
-F -- attempt to follow vforks, -h -- print help message
-i -- print instruction pointer at time of syscall
-q -- suppress messages about attaching, detaching, etc.
-r -- print relative timestamp, -t -- absolute timestamp, -tt -- with usecs
-T -- print time spent in each syscall, -V -- print version
-v -- verbose mode: print unabbreviated argv, stat, termio[s], etc. args
-x -- print non-ascii strings in hex, -xx -- print all strings in hex
-a column -- alignment COLUMN for printing syscall results (default 40)
-e expr -- a qualifying expression: option=[!]all or option=[!]val1[,val2]...
options: trace, abbrev, verbose, raw, signal, read, or write
-o file -- send trace output to FILE instead of stderr
-O overhead -- set overhead for tracing syscalls to OVERHEAD usecs
-p pid -- trace process with process id PID, may be repeated
-D -- run tracer process as a detached grandchild, not as parent
-s strsize -- limit length of print strings to STRSIZE chars (default 32)
-S sortby -- sort syscall counts by: time, calls, name, nothing (default time)
-u username -- run command as username handling setuid and/or setgid
-E var=val -- put var=val in the environment for command
-E var -- remove var from the environment for command
在诸多参数中,比较常用的有如下几个:
ü -t参数:可以显示调用操作系统过程的时间。这个在诊断中很重要,如果发现某一个执行过程时间消耗大,就可以定位到是什么问题;
ü -p参数:strace除了可以从进程启动时候开始监控之外,还可以根据操作系统进程编号,进入一个运行中的进程进行debug;
ü -o参数:输出结果如果需要以文件方式保存,就通过-o参数指定文件名称;
ü -e参数:监控表达式。提供trace、signal、read和write等内容,反映监控方面项目。常用的参数是trace=file,集中在操作系统文件访问层面。
下面,通过两个典型应用场景,来演示一下strace应用。
3、启动程序监控
strace监控研究的一个重要范例是sqlplus,很多前辈使用这个工具来测试跟踪sqlplus启动过程访问的文件形式。启动strace监控sqlplus,就是在strace命令后直接输入sqlplus命令即可。
--启动命令
[oracle@localhost ~]$ strace -t -e trace=file sqlplus /nolog
14:21:14 execve("/u01/app/oracle/bin/sqlplus", ["sqlplus", "/nolog"], [/* 31 vars */]) = 0 –启动命令
14:21:14 access("/etc/ld.so.preload", R_OK) = -1 ENOENT (No such file or directory)
14:21:14 open("/u01/app/oracle/lib/tls/x86_64/libsqlplus.so", O_RDONLY) = -1 ENOENT (No such file or directory)
14:21:14 stat("/u01/app/oracle/lib/tls/x86_64", 0x7fff3076da50) = -1 ENOENT (No such file or directory)
14:21:14 open("/u01/app/oracle/lib/tls/libsqlplus.so", O_RDONLY) = -1 ENOENT (No such file or directory)
14:21:14 stat("/u01/app/oracle/lib/tls", 0x7fff3076da50) = -1 ENOENT (No such file or directory)
14:21:14 open("/u01/app/oracle/lib/x86_64/libsqlplus.so", O_RDONLY) = -1 ENOENT (No such file or directory)
14:21:14 stat("/u01/app/oracle/lib/x86_64", 0x7fff3076da50) = -1 ENOENT (No such file or directory)
14:21:14 open("/u01/app/oracle/lib/libsqlplus.so", O_RDONLY) = 3
14:21:14 open("/u01/app/oracle/lib/libclntsh.so.11.1", O_RDONLY) = 3
14:21:14 open("/u01/app/oracle/lib/libnnz11.so", O_RDONLY) = 3
14:21:14 open("/u01/app/oracle/lib/libdl.so.2", O_RDONLY) = -1 ENOENT (No such file or directory)
(篇幅原因,有省略……)
14:21:14 access("/etc/tnsnav.ora", F_OK) = -1 ENOENT (No such file or directory)
14:21:14 access("/u01/app/oracle/network/admin/tnsnav.ora", F_OK) = -1 ENOENT (No such file or directory)
14:21:14 open("/u01/app/oracle/oracore/zoneinfo", O_RDONLY|O_NONBLOCK|O_DIRECTORY|O_CLOEXEC) = 6
14:21:14 open("/u01/app/oracle/oracore/zoneinfo/timezlrg_1.dat", O_RDONLY) = 7
14:21:14 open("/u01/app/oracle/oracore/zoneinfo/timezlrg_2.dat", O_RDONLY) = 7
(读取各种时区数据……)
14:21:14 open("/u01/app/oracle/oracore/zoneinfo/timezone_14.dat", O_RDONLY) = 7
14:21:14 open("/u01/app/oracle/oracore/zoneinfo/timezlrg_14.dat", O_RDONLY) = 6
14:21:14 open("/usr/lib64/libnuma.so", O_RDONLY) = -1 ENOENT (No such file or directory)
14:21:14 open("/usr/lib64/libnuma.so.1", O_RDONLY) = 6
14:21:14 open("/proc/self/status", O_RDONLY) = 6
14:21:14 open("/sys/devices/system/node", O_RDONLY|O_NONBLOCK|O_DIRECTORY|O_CLOEXEC) = 6
14:21:14 open("/sys/devices/system/node/node0/meminfo", O_RDONLY) = 7
14:21:14 open("/sys/devices/system/cpu", O_RDONLY|O_NONBLOCK|O_DIRECTORY|O_CLOEXEC) = 6
14:21:14 open("/proc/self/status", O_RDONLY) = 6
14:21:14 open("/proc/sys/kernel/shmmax", O_RDONLY) = 6
14:21:14 open("/proc/meminfo", O_RDONLY) = 6
14:21:14 open("/usr/lib64/libnuma.so", O_RDONLY) = -1 ENOENT (No such file or directory)
14:21:14 open("/etc/localtime", O_RDONLY) = 6
14:21:14 open("/u01/app/oracle/rdbms/mesg/ocius.msb", O_RDONLY) = 6
SQL*Plus: Release 11.2.0.4.0 Production on Fri Jun 12 14:21:14 2015
Copyright (c) 1982, 2013, Oracle. All rights reserved.
--启动之后,读取的文件。
14:21:14 stat("login.sql", 0x7fff3076c740) = -1 ENOENT (No such file or directory)
14:21:14 stat("/u01/app/oracle/sqlplus/admin/glogin.sql", {st_mode=S_IFREG|0644, st_size=368, ...}) = 0
14:21:14 access("/u01/app/oracle/sqlplus/admin/glogin.sql", F_OK) = 0
14:21:14 statfs("/u01/app/oracle/sqlplus/admin/glogin.sql", {f_type="EXT2_SUPER_MAGIC", f_bsize=4096, f_blocks=174170552, f_bfree=147038320, f_bavail=138190973, f_files=44236800, f_ffree=44019067, f_fsid={1821877354, 763674030}, f_namelen=255, f_frsize=4096}) = 0
14:21:14 open("/u01/app/oracle/sqlplus/admin/glogin.sql", O_RDONLY) = 7
如果在启动sqlplus过程中,直接进行登录。跟踪过程中还包括了登录动作。
[oracle@localhost ~]$ strace -t -e trace=file sqlplus / as sysdba
14:29:30 execve("/u01/app/oracle/bin/sqlplus", ["sqlplus", "/", "as", "sysdba"], [/* 31 vars */]) = 0
(相同篇幅内容,略……)
SQL*Plus: Release 11.2.0.4.0 Production on Fri Jun 12 14:29:30 2015
Copyright (c) 1982, 2013, Oracle. All rights reserved.
--差异部分
14:29:30 open("/u01/app/oracle/dbs/cm_xxxdb.dat", O_RDONLY|O_DSYNC) = -1 ENOENT (No such file or directory)
14:29:30 getcwd("/home/oracle", 256) = 13
14:29:30 open("/proc/self/cmdline", O_RDONLY) = 7
14:29:30 open("/etc/nsswitch.conf", O_RDONLY) = 7
14:29:30 open("/u01/app/oracle/lib/libnss_files.so.2", O_RDONLY) = -1 ENOENT (No such file or directory)
14:29:30 open("/etc/ld.so.cache", O_RDONLY) = 7
14:29:30 open("/lib64/libnss_files.so.2", O_RDONLY) = 7
14:29:30 open("/etc/passwd", O_RDONLY|O_CLOEXEC) = 7
14:29:30 open("/u01/app/oracle/lib/libnque11.so", O_RDONLY) = 7
14:29:30 getcwd("/home/oracle", 256) = 13
14:29:30 open("/u01/app/oracle/nls/data/lx20369.nlb", O_RDONLY) = 7
14:29:30 open("/u01/app/oracle/nls/data/lx207d0.nlb", O_RDONLY) = 7
14:29:30 readlink("/proc/self/fd/0", "/dev/pts/0", 254) = 10
14:29:30 open("/etc/passwd", O_RDONLY|O_CLOEXEC) = 7
14:29:30 open("/etc/host.conf", O_RDONLY) = 7
14:29:30 open("/etc/resolv.conf", O_RDONLY) = 7
14:29:30 open("/etc/hosts", O_RDONLY|O_CLOEXEC) = 7
14:29:30 open("/etc/hostid", O_RDONLY) = -1 ENOENT (No such file or directory)
14:29:30 open("/etc/hosts", O_RDONLY|O_CLOEXEC) = 7
14:29:30 open("/u01/app/oracle/rdbms/mesg/oraus.msb", O_RDONLY) = 7
Connected to:
Oracle Database 11g Enterprise Edition Release 11.2.0.4.0 - 64bit Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
14:29:30 stat("login.sql", 0x7fff0a822b10) = -1 ENOENT (No such file or directory)
14:29:30 stat("/u01/app/oracle/sqlplus/admin/glogin.sql", {st_mode=S_IFREG|0644, st_size=368, ...}) = 0
14:29:30 access("/u01/app/oracle/sqlplus/admin/glogin.sql", F_OK) = 0
14:29:30 statfs("/u01/app/oracle/sqlplus/admin/glogin.sql", {f_type="EXT2_SUPER_MAGIC", f_bsize=4096, f_blocks=174170552, f_bfree=147038317, f_bavail=138190970, f_files=44236800, f_ffree=44019066, f_fsid={1821877354, 763674030}, f_namelen=255, f_frsize=4096}) = 0
14:29:30 open("/u01/app/oracle/sqlplus/admin/glogin.sql", O_RDONLY) = 7
4、跟踪执行进程方式
除了上面介绍的方法外,strace还可以直接跟踪监控正在执行的系统进程。下面我们来观察一下strace对Oracle后台CKPT进程的监控。
[oracle@XXX-Standby-Asm ~]$ ps -ef | grep ckpt
grid 15950 1 0 Jun05 ? 00:01:02 asm_ckpt_+ASM
oracle 18608 17848 0 17:35 pts/0 00:00:00 grep ckpt
oracle 21395 1 0 Jun08 ? 00:04:27 ora_ckpt_xxxstb
从root角度,跟踪进程编号21395的后台进程。
[root@XXX-Standby-Asm ~]# strace -t -p 21395
Process 21395 attached - interrupt to quit
17:36:42 times({tms_utime=12512, tms_stime=14273, tms_cutime=0, tms_cstime=0}) = 645608271
17:36:42 semtimedop(6062085, {{16, -1, 0}}, 1, {1, 840000000}) = -1 EAGAIN (Resource temporarily unavailable)
17:36:44 getrusage(RUSAGE_SELF, {ru_utime={125, 122978}, ru_stime={142, 730301}, ...}) = 0
17:36:44 getrusage(RUSAGE_SELF, {ru_utime={125, 122978}, ru_stime={142, 730301}, ...}) = 0
17:36:44 times({tms_utime=12512, tms_stime=14273, tms_cutime=0, tms_cstime=0}) = 645608455
17:36:44 times({tms_utime=12512, tms_stime=14273, tms_cutime=0, tms_cstime=0}) = 645608455
17:36:44 read(16, "MSA210P20?Pr200330266177177"..., 80) = 80
17:36:44 read(16, "MSA210P20?Pr200330266177177"..., 80) = 80
17:36:44 times({tms_utime=12512, tms_stime=14273, tms_cutime=0, tms_cstime=0}) = 645608455
17:36:44 semtimedop(6062085, {{16, -1, 0}}, 1, {3, 0}) = -1 EAGAIN (Resource temporarily unavailable) –三秒差异
17:36:47 getrusage(RUSAGE_SELF, {ru_utime={125, 123978}, ru_stime={142, 731301}, ...}) = 0
17:36:47 getrusage(RUSAGE_SELF, {ru_utime={125, 123978}, ru_stime={142, 731301}, ...}) = 0
17:36:47 times({tms_utime=12512, tms_stime=14273, tms_cutime=0, tms_cstime=0}) = 645608755
17:36:47 times({tms_utime=12512, tms_stime=14273, tms_cutime=0, tms_cstime=0}) = 645608755
17:36:47 read(16, "MSA210P20?Pr200330266177177"..., 80) = 80
17:36:47 read(16, "MSA210P20?Pr200330266177177"..., 80) = 80
17:36:47 read(16, "MSA210P20?Pr200330266177177"..., 80) = 80
17:36:47 read(16, "MSA210P20?Pr200330266177177"..., 80) = 80
17:36:47 read(16, "MSA210P20?Pr200330266177177"..., 80) = 80
17:36:47 read(16, "MSA210P20?Pr200330266177177"..., 80) = 80
17:36:47 times({tms_utime=12512, tms_stime=14273, tms_cutime=0, tms_cstime=0}) = 645608756
17:36:47 times({tms_utime=12512, tms_stime=14273, tms_cutime=0, tms_cstime=0}) = 645608756
17:36:47 times({tms_utime=12512, tms_stime=14273, tms_cutime=0, tms_cstime=0}) = 645608756
17:36:47 times({tms_utime=12512, tms_stime=14273, tms_cutime=0, tms_cstime=0}) = 645608756
17:36:47 times({tms_utime=12512, tms_stime=14273, tms_cutime=0, tms_cstime=0}) = 645608756
17:36:47 semtimedop(6062085, {{16, -1, 0}}, 1, {3, 0}^C
Process 21395 detached
从里面,我们可以看到ckpt的动作方式。
5、结论
Strace是Linux环境下的一个非常有用的工具。用好这个工具,可以让我们更加接近内核细节,丰富知识能力。
: