Chinaunix首页 | 论坛 | 博客
  • 博客访问: 3541260
  • 博文数量: 715
  • 博客积分: 1860
  • 博客等级: 上尉
  • 技术积分: 7745
  • 用 户 组: 普通用户
  • 注册时间: 2008-04-07 08:51
个人简介

偶尔有空上来看看

文章分类

全部博文(715)

文章存档

2023年(75)

2022年(134)

2021年(238)

2020年(115)

2019年(11)

2018年(9)

2017年(9)

2016年(17)

2015年(7)

2014年(4)

2013年(1)

2012年(11)

2011年(27)

2010年(35)

2009年(11)

2008年(11)

最近访客

分类: Oracle

2022-09-17 09:33:26

补丁不生效例子

aix 7202 11.2.0.4 rac 180417 PSU
实例2 例行重启,实例1发生了意外重启

实例1告警日志
   
  1. Fri Sep 16 00:34:57 2022
  2.  Set master node info
  3.  Submitted all remote-enqueue requests
  4.  Dwn-cvts replayed, VALBLKs dubious
  5.  All grantable enqueues granted
  6. Fri Sep 16 00:34:59 2022
  7. minact-scn: Master returning as live inst:2 has inc# mismatch instinc:0 cur:8 errcnt:0
  8. Fri Sep 16 00:36:16 2022
  9. Errors in file /u01/app/oracle/diag/rdbms/orcl/orcl1/trace/orcl1_lms2_38732932.trc (incident=1488512):
  10. ORA-04030: out of process memory when trying to allocate 195248 bytes (kjc pga smbuf,KJC PGA send buffer)
  11. Incident details in: /u01/app/oracle/diag/rdbms/orcl/orcl1/incident/incdir_1488512/orcl1_lms2_38732932_i1488512.trc
  12. Use ADRCI or Support Workbench to package the incident.
  13. See Note 411.1 at My Oracle Support for error and packaging details.
  14. Fri Sep 16 00:36:17 2022
  15. Errors in file /u01/app/oracle/diag/rdbms/orcl/orcl1/trace/orcl1_lms0_19269850.trc (incident=1488496):
  16. ORA-04030: out of process memory when trying to allocate 195248 bytes (kjc pga smbuf,KJC PGA send buffer)
  17. Incident details in: /u01/app/oracle/diag/rdbms/orcl/orcl1/incident/incdir_1488496/orcl1_lms0_19269850_i1488496.trc
  18. Use ADRCI or Support Workbench to package the incident.
  19. See Note 411.1 at My Oracle Support for error and packaging details.
  20. Fri Sep 16 00:36:18 2022
  21. Errors in file /u01/app/oracle/diag/rdbms/orcl/orcl1/trace/orcl1_lms1_5505754.trc (incident=1488504):
  22. ORA-04030: out of process memory when trying to allocate 195248 bytes (kjc pga smbuf,KJC PGA send buffer)
  23. Incident details in: /u01/app/oracle/diag/rdbms/orcl/orcl1/incident/incdir_1488504/orcl1_lms1_5505754_i1488504.trc
  24. Use ADRCI or Support Workbench to package the incident.
  25. See Note 411.1 at My Oracle Support for error and packaging details.
  26. Fri Sep 16 00:36:25 2022
  27. Dumping diagnostic data in directory=[cdmp_20220916003625], requested by (instance=1, osid=19269850 (LMS0)), summary=[incident=1488496].
  28. Errors in file /u01/app/oracle/diag/rdbms/orcl/orcl1/trace/orcl1_lms0_19269850.trc (incident=1488497):
  29. ORA-04030: out of process memory when trying to allocate 195248 bytes (kjc pga smbuf,KJC PGA send buffer)
  30. Incident details in: /u01/app/oracle/diag/rdbms/orcl/orcl1/incident/incdir_1488497/orcl1_lms0_19269850_i1488497.trc
  31. Use ADRCI or Support Workbench to package the incident.
  32. See Note 411.1 at My Oracle Support for error and packaging details.
  33. Errors in file /u01/app/oracle/diag/rdbms/orcl/orcl1/trace/orcl1_lms2_38732932.trc (incident=1488513):
  34. ORA-04030: out of process memory when trying to allocate 195248 bytes (kjc pga smbuf,KJC PGA send buffer)
  35. Incident details in: /u01/app/oracle/diag/rdbms/orcl/orcl1/incident/incdir_1488513/orcl1_lms2_38732932_i1488513.trc
  36. Use ADRCI or Support Workbench to package the incident.
  37. See Note 411.1 at My Oracle Support for error and packaging details.
  38. Fri Sep 16 00:36:27 2022
  39. Sweep [inc][1488513]: completed
  40. Sweep [inc][1488512]: completed
  41. Sweep [inc][1488504]: completed
  42. Sweep [inc][1488497]: completed
  43. Errors in file /u01/app/oracle/diag/rdbms/orcl/orcl1/trace/orcl1_lms1_5505754.trc (incident=1488505):
  44. ORA-04030: out of process memory when trying to allocate 195248 bytes (kjc pga smbuf,KJC PGA send buffer)
  45. Use ADRCI or Support Workbench to package the incident.
  46. See Note 411.1 at My Oracle Support for error and packaging details.
  47. Errors in file /u01/app/oracle/diag/rdbms/orcl/orcl1/trace/orcl1_lms1_5505754.trc (incident=1488506):
  48. ORA-04030: out of process memory when trying to allocate 195248 bytes (kjc pga smbuf,KJC PGA send buffer)
  49. Use ADRCI or Support Workbench to package the incident.
  50. See Note 411.1 at My Oracle Support for error and packaging details.
  51. Errors in file /u01/app/oracle/diag/rdbms/orcl/orcl1/trace/orcl1_lms1_5505754.trc (incident=1488507):
lms2、lms0、lms1三个进程都突然发生ORA-4030异常,内存不足,看相关trc文件

  1. Dump file /u01/app/oracle/diag/rdbms/orcl/orcl1/incident/incdir_1488512/orcl1_lms2_38732932_i1488512.trc
  2. Oracle Database 11g Enterprise Edition Release 11.2.0.4.0 - 64bit Production
  3. With the Partitioning, Real Application Clusters, Automatic Storage Management, OLAP,
  4. Data Mining and Real Application Testing options
  5. ORACLE_HOME = /u01/app/oracle/product/11.2.0/db_1
  6. System name:    AIX
  7. Node name:    db1
  8. Release:    2
  9. Version:    7
  10. Machine:    00FB76794C00
  11. Instance name: orcl1
  12. Redo thread mounted by this instance: 1
  13. Oracle process number: 15
  14. Unix process pid: 38732932, image: oracle@db1 (LMS2)


  15. *** 2022-09-16 00:36:16.929
  16. *** SESSION ID:(1892.1) 2022-09-16 00:36:16.929
  17. *** CLIENT ID:() 2022-09-16 00:36:16.929
  18. *** SERVICE NAME:(SYS$BACKGROUND) 2022-09-16 00:36:16.929
  19. *** MODULE NAME:() 2022-09-16 00:36:16.929
  20. *** ACTION NAME:() 2022-09-16 00:36:16.929
  21.  
  22. Dump continued from file: /u01/app/oracle/diag/rdbms/orcl/orcl1/trace/orcl1_lms2_38732932.trc
  23. ORA-04030: out of process memory when trying to allocate 195248 bytes (kjc pga smbuf,KJC PGA send buffer)

  24. ========= Dump for incident 1488512 (ORA 4030) ========
  25. ----- Beginning of Customized Incident Dump(s) -----
  26. =======================================
  27. TOP 10 MEMORY USES FOR THIS PROCESS
  28. ---------------------------------------

  29. *** 2022-09-16 00:36:22.939
  30. 99% 31 GB, 172610 chunks: "KJC PGA send buffer "       <--------------这里很异常
  31.          kjc pga smbuf ds=110733230 dsprt=110005120
  32.  1% 213 MB, 172724 chunks: "free memory "
  33.          pga heap ds=110005120 dsprt=0
  34.  0% 149 MB, 145 chunks: "permanent memory "
  35.          pga heap ds=110005120 dsprt=0
  36.  0% 75 KB, 3 chunks: "permanent memory "
  37.          session heap ds=1109d66e0 dsprt=11011f5d8
  38.  0% 54 KB, 1 chunk : "free memory "
  39.          top call heap ds=11011f3b8 dsprt=0
  40.  0% 38 KB, 1 chunk : "Fixed Uga "
  41.          Fixed UGA heap ds=1108bad28 dsprt=110005120
  42.  0% 20 KB, 2 chunks: "free memory "
  43.          session heap ds=1109d66e0 dsprt=11011f5d8
  44.  0% 16 KB, 7 chunks: "permanent memory "
  45.          diag pga ds=1107324d0 dsprt=110005120
  46.  0% 16 KB, 2 chunks: "dbgeInitProcessCtx:InvCtx "
  47.          diag pga ds=1107324d0 dsprt=110005120
  48.  0% 16 KB, 1 chunk : "kgh stack "
  49.          pga heap ds=110005120 dsprt=0
  50.  
  51. =======================================
  52. PRIVATE MEMORY SUMMARY FOR THIS PROCESS
  53. ---------------------------------------
  54. ******************************************************
  55. PRIVATE HEAP SUMMARY DUMP
  56. 32 GB total:
  57.     31 GB commented, 149 MB permanent
  58.    213 MB free (0 KB in empty extents),
  59.       32 GB, 1 heap: "kjc pga smbuf " 212 MB free held
  60. ------------------------------------------------------
  61. Summary of subheaps at depth 1
  62. 31 GB total:
  63.     31 GB commented, 98 KB permanent
  64.     31 KB free (0 KB in empty extents),
  65.       31 GB, 172610 chunks: "KJC PGA send buffer "

  66. *** 2022-09-16 00:36:23.380
  67.  
  68. =========================================
  69. REAL-FREE ALLOCATOR DUMP FOR THIS PROCESS
  70. -----------------------------------------
  71.  
  72. Dump of Real-Free Memory Allocator Heap [0x1109a4810]
  73. mag=0xfefe0001 flg=0x5000003 fds=0x0 blksz=65536
  74. blkdstbl=0x1109a4820, iniblk=521216 maxblk=524288 numsegs=320
  75. In-use num=172723 siz=4028628992, Freeable num=1 siz=65536, Free num=310 siz=20316160
  76.  
  77. ==========================================
  78. INSTANCE-WIDE PRIVATE MEMORY USAGE SUMMARY
  79. ------------------------------------------
  80.  
  81. Dumping Work Area Table (level=1)
  82. =====================================
  83.  
  84.   Global SGA Info
  85.   ---------------
  86.  
  87.     global target: 115507 MB
  88.     auto target: 11333 MB
  89.     max pga: 2048 MB
  90.     pga limit: 4096 MB
  91.     pga limit known: 0
  92.     pga limit errors: 0
  93.  
  94.     pga inuse: 102955 MB
  95.     pga alloc: 105793 MB
  96.     pga freeable: 742 MB
  97.     pga freed: 43967956 MB
  98.     pga to free: 0
  99.     broker request: 0
  100.  
  101.     pga auto: 41 MB
  102.     pga manual: 0 MB
  103.  
  104.     pga alloc (max): 105793 MB
  105.     pga auto (max): 5444 MB
  106.     pga manual (max): 1 MB
  107.  
  108.     # workareas : 43
  109.     # workareas(max): 224
  110.  
  111.  
  112.  
  113. ================================
  114. PER-PROCESS PRIVATE MEMORY USAGE
  115. --------------------------------
  116.  
  117. Private memory usage per Oracle process
  118.  
  119. -------------------------
  120. Top 10 processes:
  121. -------------------------
  122. (percentage is of 103 GB total allocated memory)
  123. 31% pid 13: 32 GB used of 32 GB allocated (128 KB freeable)
  124. 31% pid 14: 32 GB used of 32 GB allocated
  125. 31% pid 15: 32 GB used of 32 GB allocated (64 KB freeable) <= CURRENT PROC
  126.  0% pid 12: 14 MB used of 113 MB allocated (98 MB freeable)
  127.  0% pid 10: 45 MB used of 56 MB allocated (7680 KB freeable)
  128.  0% pid 86: 42 MB used of 45 MB allocated
  129.  0% pid 87: 42 MB used of 45 MB allocated
  130.  0% pid 89: 42 MB used of 45 MB allocated
  131.  0% pid 973: 42 MB used of 43 MB allocated (1024 KB freeable)
  132.  0% pid 1029: 42 MB used of 43 MB allocated (1024 KB freeable)
  133.  
  134. -------------------------
  135. All processes:
  136. -------------------------
  137. (session detail when over 5290 MB allocated)
  138.  
  139. pid 2: 2119 KB used of 2350 KB allocated
  140. pid 3: 1060 KB used of 1182 KB allocated
  141. pid 4: 1056 KB used of 1182 KB allocated
  142. pid 5: 1066 KB used of 1182 KB allocated
  143. pid 6: 6161 KB used of 6494 KB allocated (128 KB freeable)
  144. pid 7: 2121 KB used of 2718 KB allocated (320 KB freeable)
  145. ...

三个进程每个都32G,搜索MOS,发现有个bug很像

检查当前的补丁,opatch lsinv|grep 18316692 已经包含了此修复,但是LMS的pga还是很大,看来没有解决。

怎么办呢?

监控,发现内存占用太大就重启实例(LMS进程杀不得)

  1. select pid,spid,program,pga_used_mem,pga_alloc_mem,pga_freeable_mem,pga_max_mem from v$process where program like '%LMS%';

lms进程的工作机制



参考:
  1. https://orainternals.files.wordpress.com/2008/03/riyaj_battle_of_nodes_rac_perf_myths_ppt.pdf


阅读(1130) | 评论(0) | 转发(0) |
给主人留下些什么吧!~~