crs-0184错误，请帮帮

mason003

环境： aix6.1  oracle 10.2.0.4
状况：
1.其中一个节点执行crs_stat -t 报错 CRS-0184: Cannot communicate with the CRS daemon.
2.上述节点ps -ef|grep ora_ 数据库相关进程都在
3.上述节点可以sqlplus 登陆
4.上树节点root执行
crsctl start crs 提示
Attempting to start CRS stack
The CRS stack will be started shortly
然后oracle执行crs_stat -t提示 CRS-0184: Cannot communicate with the CRS daemon

请帮帮看看这个问题，如果需要提供日志，请告诉我提供哪些日志

Maclean Liu(刘相兵

把问题节点上的

CRS logs are in $ORA_CRS_HOME/log/<hostname>/crsd/crsd.log
CSS logs are in $ORA_CRS_HOME/log/<hostname>/cssd/ocssd.log

crsctl check crs
crsctl check css

mason003

# /oracle/crs/bin/crsctl check css
Failure 1 contacting CSS daemon
# /oracle/crs/bin/crsctl check crs
Failure 1 contacting CSS daemon
Cannot communicate with CRS
Cannot communicate with EVM
crsd.log.txt (5.23 MB, 下载次数: 1011)

2012-5-22 09:15 上传

点击文件名下载附件

ocssd.log.txt (276 KB, 下载次数: 1020)

2012-5-22 09:15 上传

点击文件名下载附件

ricky

你在root用户下面，开启一下crs，这个就是crs没有打开，或是遇到什么问题dang了。如果可以重启的话，我建议先正常关闭，在按照步骤来启动。

mason003

楼上的哥们儿， 你的意思是把整个服务器重启了？

miloluo

1. 
   
2. crsd.log shows error 'Failed to spawn a thread for UI connection. error=-11' [ID 1389408.1]       
   
3. 
   
4. 修改时间:2011-12-30类型:PROBLEM状态:PUBLISHED优先级:3       
   
5. 
   
6. 
   
7. In this Document
   
8.   Symptoms
   
9.   Cause
   
10.   Solution
    
11.   References
    
12. 
    
13. Applies to:
    
14. Oracle Server - Enterprise Edition - Version: 10.2.0.1 to 11.1.0.7 - Release: 10.2 to 11.1
    
15. Generic Linux
    
16. Generic UNIX
    
17. Symptoms
    
18. <CRS_HOME>/log/<nodename>/crsd/crsd.log shows below error:
    
19. 
    
20. 2011-12-27 11:41:13.404: [ CRSMAIN][10287] Failed to spawn a thread for UI connection. error=-11
    
21. 
    
22. 
    
23. "ps -ef|grep crsd.bin" shows crsd.bin is up and running, but crs_stat shows the following error:
    
24. 
    
25. crs_stat -t
    
26. CRS-0184: Cannot communicate with the CRS daemon.
    
27. 
    
28. 
    
29. Cause
    
30. Refer to Solution Section.
    
31. 
    
32. Solution
    
33. 1. Incorrect stack setting for root user:
    
34. 
    
35. #ulimit -a
    
36. time(seconds) unlimited
    
37. file(blocks) 2097151
    
38. data(kbytes) 131072
    
39. stack(kbytes) 32768<==========
    
40. memory(kbytes) 32768
    
41. coredump(blocks) 2097151
    
42. nofiles(descriptors) 2000
    
43. threads(per process) unlimited
    
44. processes(per user) unlimited
    
45. 
    
46. The issue happens more often on AIX, and the stack size should be unlimited.
    
47. 
    
48. 
    
49. 
    
50. 2. Unpublished bug 6117687 - CRS slow memory leak from repeated "crsctl check crs"
    
51. 
    
52. The bug is fixed in 10.2.0.4, 11.1.0.6 and above

复制代码

可以查查这里面提到的内容，试试看。

Maclean Liu(刘相兵

ODM DATA:

1. [    CSSD]2011-11-16 10:34:00.153 [2829] >TRACE:   clssgmHandleDBDone(): src/dest (1/65535) size(72) incarn 6
   
2. [    CSSD]CLSS-3000: reconfiguration successful, incarnation 6 with 2 nodes
   
3. 
   
4. [    CSSD]CLSS-3001: local node number 2, master node number 1
   
5. 
   
6. [    CSSD]2011-11-16 10:34:00.154 [3857] >TRACE:   clssgmReconfigThread:  completed for reconfig(6), with status(1)
   
7. [    CSSD]2011-11-16 10:34:00.582 [2829] >TRACE:   clssgmCommonAddMember: clsomon joined (2/0x1000000/#CSS_CLSSOMON)
   
8. [    CSSD]2012-02-21 22:34:28.368 [2572] >TRACE:   Authorization failed (111a818f0), timed out, start 22:32:28.368, duration 120000
   
9. 
   
10. [    CSSD]2012-02-21 22:34:28.368 [2572] >TRACE:   Authorization prepare time: 1 ms
    
11. 
    
12. [    CSSD]2012-02-21 22:34:29.608 [2572] >TRACE:   Authorization failed (111aa6550), timed out, start 22:32:29.608, duration 120000
    
13. 
    
14. [    CSSD]2012-02-21 22:34:29.608 [2572] >TRACE:   Authorization prepare time: 1 ms
    
15. 
    
16. [    CSSD]2012-02-21 22:35:30.176 [2572] >TRACE:   Authorization failed (111ab6950), timed out, start 22:33:30.174, duration 120001
    
17. 
    
18. [    CSSD]2012-02-21 22:35:30.176 [2572] >TRACE:   Authorization prepare time: 0 ms
    
19. 
    
20. [    CSSD]2012-02-21 22:36:30.665 [2572] >TRACE:   Authorization failed (111a818f0), timed out, start 22:34:30.664, duration 120000
    
21. 
    
22. [    CSSD]2012-02-21 22:36:30.665 [2572] >TRACE:   Authorization prepare time: 0 ms
    
23. 
    
24. [    CSSD]2012-02-21 22:37:31.153 [2572] >TRACE:   Authorization failed (111aa6550), timed out, start 22:35:31.153, duration 120000
    
25. 
    
26. [    CSSD]2012-02-21 22:37:31.153 [2572] >TRACE:   Authorization prepare time: 1 ms
    
27. 
    
28. [    CSSD]2012-02-21 22:37:49.164 [2572] >TRACE:   Authentication OSD error, op: scls_auth_response_prepare
    
29. loc: mkdir
    
30. info: failed to make dir /oracle/crs/css/auth/A3861086, No space left on device
    
31. dep: 28
    
32. [    CSSD]2012-02-21 22:37:49.359 [2572] >TRACE:   Authentication OSD error, op: scls_auth_response_prepare
    
33. loc: mkdir
    
34. info: failed to make dir /ora

复制代码

Maclean Liu(刘相兵

就日志 论日志 而言， 你上传的node2的日志 说明：

这套CRS 应当是在2011-05-22 安装的，并升级到 10.2.0.4

最近一次的成功reconfiguration  是在 2011-11-16 10:34:00.582

之后在2012-02-21 css被再次启动过 可能源于 重启


但是出现了 一下错误

[    CSSD]2012-02-21 22:37:49.164 [2572] >TRACE:   Authentication OSD error, op: scls_auth_response_prepare
loc: mkdir
info: failed to make dir /oracle/crs/css/auth/A3861086, No space left on device



这应当是由于  /oracle/crs/css/auth/A3861086 所在mount点 没有剩余磁盘空间所致

crsd.log 日志 也是现实相似的信息

2012-01-17 23:24:03.955: [  CRSRES][12351]32Start of `ora.portaldb02.LISTENER_PORTALDB02.lsnr` on member `portaldb02` succeeded.
2012-01-17 23:24:08.544: [  OCRUTL][5410]u_freem: mem passed is null
2012-01-24 01:45:07.030: [ COMMCRS][3611]Authorization failed, network error

ricky

原来是磁盘空间满了啊！这个问题怎么没有人检查吗？哎~~不该啊