复杂环境带来的坏块

monkeybron

各位大佬好：
  附件是报错的trc 文件

我说下环境：
                        虚拟环境   hyper-v 群集  windows 2013 操作系统



                          虚拟机 os  redhat6.2   11g rac     版本11.2.0.1.0 ,2台节点分别在2个  hyper-v 群集上面


                         alert 出现：

1. Sat Nov 15 22:25:26 2014
   
2. Hex dump of (file 14, block 459853) in trace file /u01/app/oracle/diag/rdbms/cordysdb/cordysdb1/trace/cordysdb1_ora_51329.trc
   
3. Corrupt block relative dba: 0x0387044d (file 14, block 459853)
   
4. Bad header found during backing up datafile
   
5. Data in bad block:
   
6. type: 40 format: 2 rdba: 0x038703cd
   
7. last change scn: 0x091d.4f968c3f seq: 0x2 flg: 0x04
   
8. spare1: 0x0 spare2: 0x0 spare3: 0x3
   
9. consistency value in tail: 0x8c3f2802
   
10. check value in block header: 0xee4b
    
11. computed block checksum: 0x0
    
12. Reread of blocknum=459853, file=+CORDYS_DATA01/cordysdb/datafile/dts_cordysbpm.271.827227815. found valid data
    
13. Hex dump of (file 14, block 459854) in trace file /u01/app/oracle/diag/rdbms/cordysdb/cordysdb1/trace/cordysdb1_ora_51329.trc
    
14. Corrupt block relative dba: 0x0387044e (file 14, block 459854)
    
15. Bad header found during backing up datafile
    
16. Data in bad block:
    
17. type: 40 format: 2 rdba: 0x038703ce
    
18. last change scn: 0x091d.4fa68ab2 seq: 0x1 flg: 0x04
    
19. spare1: 0x0 spare2: 0x0 spare3: 0x3
    
20. consistency value in tail: 0x8ab22801
    
21. check value in block header: 0xca3d
    
22. computed block checksum: 0x0
    
23. Reread of blocknum=459854, file=+CORDYS_DATA01/cordysdb/datafile/dts_cordysbpm.271.827227815. found valid data
    
24. Hex dump of (file 14, block 459855) in trace file /u01/app/oracle/diag/rdbms/cordysdb/cordysdb1/trace/cordysdb1_ora_51329.trc
    
25. Corrupt block relative dba: 0x0387044f (file 14, block 459855)
    
26. Bad header found during backing up datafile
    
27. Data in bad block:
    
28. type: 40 format: 2 rdba: 0x038703cf
    
29. last change scn: 0x0000.35e51ab6 seq: 0x2 flg: 0x04
    
30. spare1: 0x0 spare2: 0x0 spare3: 0x3
    
31. consistency value in tail: 0x1ab62802
    
32. check value in block header: 0x28ce
    
33. computed block checksum: 0x0
    
34. Reread of blocknum=459855, file=+CORDYS_DATA01/cordysdb/datafile/dts_cordysbpm.271.827227815. found valid data
    
35. Sat Nov 15 23:37:37 2014
    
36. Hex dump of (file 14, block 2194939) in trace file /u01/app/oracle/diag/rdbms/cordysdb/cordysdb1/trace/cordysdb1_ora_51329.trc
    
37. Corrupt block relative dba: 0x03a17dfb (file 14, block 2194939)
    
38. Bad header found during backing up datafile
    
39. Data in bad block:
    
40. type: 40 format: 2 rdba: 0x03a17d7b
    
41. last change scn: 0x091d.3b0b0a74 seq: 0x1 flg: 0x04
    
42. spare1: 0x0 spare2: 0x0 spare3: 0x3
    
43. consistency value in tail: 0x0a742801
    
44. check value in block header: 0x3a02
    
45. computed block checksum: 0x0
    
46. Reread of blocknum=2194939, file=+CORDYS_DATA01/cordysdb/datafile/dts_cordysbpm.271.827227815. found valid data
    
47. Hex dump of (file 14, block 2194940) in trace file /u01/app/oracle/diag/rdbms/cordysdb/cordysdb1/trace/cordysdb1_ora_51329.trc
    
48. Corrupt block relative dba: 0x03a17dfc (file 14, block 2194940)
    
49. Bad header found during backing up datafile
    
50. Data in bad block:
    
51. type: 40 format: 2 rdba: 0x03a17d7c
    
52. last change scn: 0x091d.37215e68 seq: 0x2 flg: 0x04
    
53. spare1: 0x0 spare2: 0x0 spare3: 0x3
    
54. consistency value in tail: 0x5e682802
    
55. check value in block header: 0x8323
    
56. computed block checksum: 0x0
    
57. Reread of blocknum=2194940, file=+CORDYS_DATA01/cordysdb/datafile/dts_cordysbpm.271.827227815. found valid data
    
58. Hex dump of (file 14, block 2194941) in trace file /u01/app/oracle/diag/rdbms/cordysdb/cordysdb1/trace/cordysdb1_ora_51329.trc
    
59. Corrupt block relative dba: 0x03a17dfd (file 14, block 2194941)
    
60. Bad header found during backing up datafile
    
61. Data in bad block:
    
62. type: 40 format: 2 rdba: 0x03a17d7d
    
63. last change scn: 0x091d.4048e976 seq: 0x1 flg: 0x04
    
64. spare1: 0x0 spare2: 0x0 spare3: 0x3
    
65. consistency value in tail: 0xe9762801
    
66. check value in block header: 0x718f
    
67. computed block checksum: 0x0
    
68. Reread of blocknum=2194941, file=+CORDYS_DATA01/cordysdb/datafile/dts_cordysbpm.271.827227815. found valid data
    
69. Hex dump of (file 14, block 2194942) in trace file /u01/app/oracle/diag/rdbms/cordysdb/cordysdb1/trace/cordysdb1_ora_51329.trc
    
70. Corrupt block relative dba: 0x03a17dfe (file 14, block 2194942)
    
71. Bad header found during backing up datafile
    
72. Data in bad block:
    
73. type: 40 format: 2 rdba: 0x03a17d7e
    
74. last change scn: 0x091d.3b432042 seq: 0x1 flg: 0x04
    
75. spare1: 0x0 spare2: 0x0 spare3: 0x3
    
76. consistency value in tail: 0x20422801
    
77. check value in block header: 0xd9c2
    
78. computed block checksum: 0x0
    
79. Reread of blocknum=2194942, file=+CORDYS_DATA01/cordysdb/datafile/dts_cordysbpm.271.827227815. found valid data
    
80. Hex dump of (file 14, block 2194943) in trace file /u01/app/oracle/diag/rdbms/cordysdb/cordysdb1/trace/cordysdb1_ora_51329.trc
    
81. Corrupt block relative dba: 0x03a17dff (file 14, block 2194943)
    
82. Bad header found during backing up datafile
    
83. Data in bad block:
    
84. type: 40 format: 2 rdba: 0x03a17d7f
    
85. last change scn: 0x091d.361488eb seq: 0x2 flg: 0x04
    
86. spare1: 0x0 spare2: 0x0 spare3: 0x3
    
87. consistency value in tail: 0x88eb2802
    
88. check value in block header: 0xc6a4
    
89. computed block checksum: 0x0
    
90. Reread of blocknum=2194943, file=+CORDYS_DATA01/cordysdb/datafile/dts_cordysbpm.271.827227815. found valid data
    
91. Sun Nov 16 00:00:01 2014
    
92. Trace dumping is performing id=[cdmp_20141116000000]

复制代码

Maclean Liu(刘相兵

Reread of blocknum=2194941, file=+CORDYS_DATA01/cordysdb/datafile/dts_cordysbpm.271.827227815. found valid data


再次读取发现 块是可用的，这可能和虚拟环境有关。

可以参考这里 http://www.askmaclean.com/archiv ... ng-up-datafile.html

就是说rman在备份期间读取到该数据块，并在初次读取时发现该块断裂了(Fractured)，但Fractured并不代表块就真的corrupted了；从告警日志看rman在初次读取发现该块Fractured后，又再次读取时发现数据块已经恢复正常(found valid data)。所以上述告警日志并代表所列出的数据块存在讹误，很有可能是该数据块所在数据文件在备份期间发生了剧烈的IO操作，当rman读取到该数据块时可能存储正在对其进行写的操作，所以rman在第一次读取时认为该快断裂了(Fractured)；之后rman对该块进行reread发现”断裂”现象已不存在，而”Corrupt block”仅仅是一种假象；针对上述问题可以对表或索引进行进一步的analyze..validate操作以确保不存在坏块。
同时上述”Corrupt block误报”现象极有可能是因为在Rman备份期间个别数据文件的IO过于活跃所致(如频繁的dml操作)，建议在磁盘活跃度低的时间段运行rman备份工作。