opiodr aborting process unknown ospid (58064936) as a result of ORA-609
1、数据库版本 11.2.0.1操作系统 aix 6.1
topas
Forks 0Igets 0
NetworkKBPS I-PackO-Pack KB-InKB-OutExecs 0Namei 7
Total 246.6 739.7 763.6 176.2 70.4Runqueue 0.0Dirblk 0
Waitqueue 1.5
Disk Busy% KBPS TPS KB-Read KB-Writ MEMORY
Total 33.0 3416.9 183.0 175.93241.0PAGING Real,MB 31488
Faults 44% Comp 75
FileSystem KBPS TPS KB-Read KB-WritSteals 0% Noncomp 6
Total 2.9K146.9 3.0 2.9K PgspIn 0% Client 6
PgspOut 0
Name PIDCPU%PgSp Owner PageIn 43PAGING SPACE
oracle 9306128 0.116.9 oracle PageOut 810Size,MB 24576
oracle 57737308 0.110.8 oracle Sios 837% Used 0
oracle 32964656 0.110.8 oracle % Free 100
oracle 12517380 0.117.0 oracle NFS (calls/sec)
oracle 58065024 0.110.8 oracle SerV2 0WPAR Activ 0
oracle 22610142 0.117.0 oracle CliV2 0WPAR Total 0
oracle 8650776 0.111.1 oracle SerV3 0Press: "h"-help
oracle 16056444 0.016.9 oracle CliV3 0 "q"-quit
oracle 131378 0.016.9 oracle
aioserve 31719624 0.0 0.4 oracle
aioserve 49021146 0.0 0.4 oracle
oracle 7602328 0.015.9 oracle
oracle 4849860 0.024.7 oracle
aioserve 66191588 0.0 0.4 oracle
xmgc 917532 0.0 0.4 root
oracle 31523012 0.016.7 oracle
oracle 59637882 0.010.1 oracle
topas 27852852 0.0 4.6 oracle
oracle 7733292 0.015.9 oracle
oracle 7995484 0.015.9 oracle
目前数据库很空闲，但是vmstat 还是有很多进程被阻塞，
vmstats 1 100
System configuration: lcpu=32 mem=31488MB
kthr memory page faults cpu
----- ----------- ------------------------ ------------ -----------
rb avm frerepipofr srcyin sycs us sy id wa
01 6072640 1472707 0 2 0 212210 0 1017 14895 374211 945
01 6072640 1472326 0 0 0 0 0 0 806 12601 322010 945
01 6072640 1472183 0 0 0 0 0 0 1485 26108 532721 944
11 6072640 1472057 0 0 0 0 0 0 1464 21511 498111 953
02 6072640 1471295 0 0 0 0 0 0 869 11497 322010 954
01 6072640 1470986 0 0 0 0 0 0 1034 14155 379910 954
01 6072640 1470876 0 0 0 0 0 0 1177 17224 415110 954
02 6072640 1470049 0 0 0 0 0 0 976 14096 348310 945
01 6072717 1469770 0 0 096 96 0 1391 21264 497011 944
01 6072717 1469607 0 0 0 0 0 0 1307 20216 487211 944
02 6072696 1474044 0 0 0 5198 5232 0 1204 23574 454921 934
01 6072698 1473742 0 0 0 0 0 0 754 11273 308410 936
00 6072698 1473552 0 0 0 0 0 0 1524 26218 534221 944
01 6072621 1472761 0 0 0 0 0 0 1196 19460 439111 945
02 6072621 1472484 0 0 0 0 0 0 895 14167 370510 935
01 6072621 1472383 0 0 0 0 0 0 1066 13846 375210 954
如果连接进来的进程日志回报
Errors in file /data/u01/app/oracle/diag/rdbms/newbos/NEWBOS/trace/NEWBOS_ora_58064936.trc:
ORA-00609: could not attach to incoming connection
ORA-12537: TNS:connection closed
opiodr aborting process unknown ospid (58064936) as a result of ORA-609
Fri Jul 04 12:14:59 2014
Trace dumping is performing id=
Fri Jul 04 12:18:21 2014
Errors in file /data/u01/app/oracle/diag/rdbms/newbos/NEWBOS/trace/NEWBOS_ora_36372572.trc:
ORA-00609: could not attach to incoming connection
ORA-12537: TNS:connection closed
opiodr aborting process unknown ospid (36372572) as a result of ORA-609
Fri Jul 04 12:18:23 2014
Trace dumping is performing id=
Fri Jul 04 12:18:38 2014
这里有几个问题：
Kthr
kernel thread state(内核线程状态,表示每秒钟在采样间隔时间上对各种队列的内核线程数求得的平均值)
r: 取样期间可运行的内核线程的平均数，包括正在运行的线程和指准备运行但尚在等待运行的进程。
1.)如果在processes中运行的序列(process r)是连续的大于在系统中的CPU的个数表示系统现在运行比较慢,有多数的进程等待CPU。
2.)如果r的输出数大于系统中可用CPU个数的4倍的话,则系统面临着CPU短缺的问题,或者是CPU的速率过低,系统中有多数的进程在等待CPU,造成系统中进程运行过慢。
3.)如果空闲时间(cpu id)持续为0并且系统时间(cpu sy)是用户时间的两倍(cpu us)系统则面临着CPU资源的短缺。
解决方法
当发生以上问题的时候请先调整应用程序对CPU的占用情况.使得应用程序能够更有效的使用CPU.同时可以考虑增加更多的CPU.一般情况下，应用程序的问题会比较大一些.比如一些sql语句不合理等等都会造成这样的现象.
b：表示每秒VMM等待队列中的内核线程平均数(等待资源或I/O)，这里参考值为2，大于2表示被阻塞列线程数目太多。
ORA-00609和ORA-12537
从11g开始，很多以前版本写入监听日志的东西，写入到了数据库的alert日志，对于发现该错误，一般分析思路：
0. 检查你的网络和主机网络相关配置
1. 检查监听日志
2. 检查sqlnet.ora
3. 对于tns的服务端和客户端如果需要可以做trace分析
2. Oracle Net Level 16 Client tracing. Add to a clients SQLNET.ORA file
DIAG_ADR_ENABLED=off # Disable ADR if version 11g
TRACE_LEVEL_CLIENT = 16 # Enable level 16 trace
TRACE_TIMESTAMP_CLIENT = ON # Set timestamp in the trace files
TRACE_DIRECTORY_CLIENT = <DIRECTORY># Control trace file location
TRACE_FILELEN_CLIENT =<n> #Control size of trace set in kilobytes eg 20480
TRACE_FILENO_CLIENT =<n> #Control number of trace files per process
If the connection model is JDBC thin, Javanet tracing of a client is required. See Document 793415.1 How to Perform the Equivalent of SQL*Net Client Tracing with Oracle JDBC Thin Driver.
If 11.2 JDBC thin client used, the following note can be used Document 1050942.1 How to Trace the Network Packets Exchanged Between JDBC and the RDBMS in Release 11.2
3. Oracle Net Level 16 Server tracing. Add to server side SQLNET.ORA file
DIAG_ADR_ENABLED=off # Disable ADR if version 11g
TRACE_LEVEL_SERVER = 16 # Enable level 16 trace
TRACE_TIMESTAMP_SERVER = ON # Set timestamp in the trace files
TRACE_DIRECTORY_SERVER = <DIRECTORY># Control trace file location
TRACE_FILELEN_SERVER =<n> #Control size of trace set in kilobytes eg 20480
TRACE_FILENO_SERVER =<n> #Control number of trace files per process
另外你还可以通过SQLNET.INBOUND_CONNECT_TIMEOUT配置来改善
具体可以参考：
Troubleshooting Guide ORA-609 : Opiodr aborting process unknown ospid (Doc ID 1121357.1)
Troubleshooting ORA-12537 / TNS-12537 TNS:Connection Closed (Doc ID 555609.1)
页:
[1]