Chinaunix

标题: HACMP的cluster.log中不断报错 [打印本页]
作者: nds521    时间: 2012-07-23 09:14
标题: HACMP的cluster.log中不断报错
本帖最后由 nds521 于 2012-07-23 09:16 编辑

HACMP总是每隔一个小时生成一条如下信息，不知该如何解决，求助帮忙看看

Jul 23 08:02:09 di01 user:err|error syslog: slp: 0660-084 [5570738] The SA failed to decode and compute received message: Parse Error (-2).
Jul 23 09:02:09 di01 user:err|error syslog: slp: 0660-065 [5570738] Impossible to parse attribute (ca-uid=file:///var/opt/tivoli/ep/runtime/agent),(am-host=),(ca-ips=192.168.10.14\2c 192.168.10.4\2c 172.16.12.13\2c 172.16.11.13),(ca-basic-port=9510),(ca-cert-port=9510),(ca-version=1.4.2.2),(os-uid=53E8CBCC19B611E195E1000000000000).
作者: zwz99999    时间: 2012-07-23 16:11
心跳有问题吧！！！！！！！！！！！！！
作者: 好运北京    时间: 2012-07-23 16:13
偶也搞不明白
作者: hello_unix    时间: 2012-07-23 22:07
同步验证一下，看看报什么错
作者: dyllove98    时间: 2012-07-24 11:34
Impossible to parse attribute (ca-uid=file:///var/opt/tivoli/ep/runtime/agent
作者: nds521    时间: 2012-07-25 09:47
本帖最后由 nds521 于 2012-07-25 11:09 编辑

回复 4# hello_unix
不好意思啊！请问怎么做同步验证呢？
同步验证之后需要做什么？
还有就是验证时会不会对系统造成影响？
我才接触小型机，对于HA的配置和使用还是不太了解，请版主多多帮助一下。
作者: InfoSVC    时间: 2012-07-25 15:09
smitty  hacmp里面有个子菜单可以做同步
同步对系统没有啥影响
只是看看同步的时候会提示啥信息
作者: yddll    时间: 2012-07-26 21:51
lslpp -l |grep -i tivoli看看
作者: zhangxuwl    时间: 2012-07-28 22:07
ha的软件是不是多装了几个没必要的东西呢。一般这三个不用装 nfs  plugin 还有一个havitor神马的。忘了怎么拼些了。
/usr/es/sbin/cluster/utilities/cltopinfo  -m 查查心跳包有没有丢。没丢说明心跳正常。或者lssrc -ls topsvcs 都可以看心跳有没有丢包。
在查查资源组，拓扑状态。没有异常就不要担心。不过就不知道这玩意会不会把/var/目录干满。导致报错，仔细观察呗。
作者: jat_15    时间: 2012-07-29 20:38
看下hacmp.out里面有没有对应时间点的事件，在进行故障excute.
作者: nds521    时间: 2012-08-01 11:26
按照大家给的建议都检查过了。
HA心跳和同步验证都没有问题，现在还是在cluster.log中每隔一小时产生一次这个信息。
作者: nds521    时间: 2012-08-01 11:30
yddll 发表于 2012-07-26 21:51
lslpp -l |grep -i tivoli看看


然后该怎么做，跟tivoli的安装有关吗？
作者: coolcat982    时间: 2013-12-25 18:28
- I asked if /var/adm was filling up...  ->
it's redirecting to envision central server,it's filling up in that
server
-- what exact version of AIX (oslevel -s )?
IV00823  SLP_SRVREG USES 100% CPU OR DUMPS CORE
IV07482  SLP_SRVREG LEAKS MEMORY
IV02032  SLP_SRVREG FILLS SYSLOG WITH DEBUG MESSAGES
IV07657  SLP_SRVREG USES LARGE AMOUNTS OF CPU BUT IS STILL RESPONSIVE
# oslevel -s
6100-06-03-1048
- He didn't know if they use IBM Systems Director...
http://www-03.ibm.com/systems/software/director/
If you are not using IBM Systems Director (which is usually the case)
you can do the following to stop the daemons and eliminate the unwanted
messages. # cp /etc/inittab /etc/inittab.bak
# vi /etc/inittab
Comment out the following two lines with a colon.
:platform_agent:2nce:/usr/bin/startsrc -s platform_agent >/dev/null
2>&1   (doesn't apply in this case)
:cimservices:2nce:/usr/bin/startsrc -s cimsys >/dev/null 2>&1
Save the file.
.
Refresh the inittab file
# telinit q
# stopsrc -s platform_agent
# stopsrc -s cimsys
# ps -ef |grep slp
If you have slp_srvreg process running, kill process id if running.
.
The above steps should prevent slp from logging to syslog.



Action Taken:
-------------
- Nick Gimben contacted me regarding the above messages, which are
related to Service Location Protocol (SLP).

From the snap:
# ps -ef | grep slp_srvreg
     UID      PID     PPID   C    STIME    TTY  TIME CMD
    root  4194436        1   0 13:01:01      -  0:09 ./slp_srvreg -D

# lssrc -a
Subsystem         Group            PID          Status
platform_agent                    3670134      active
cimsys                            3211398      active

- SLP provides a flexible and scalable framework for providing hosts
with access to information about the existence, location, and
configuration of network services.
- The SLP service is primarily used by IBM Systems Director.

- Customer is not using IBM Systems Director, he can go ahead and
disable the SLP service using the following steps:
1) # slp_srvreg -k
If the slp_srvreg server is still running after using "slp_srvreg -k",
kill it manually with a kill -9:
# ps -ef | grep slp_srvreg
# kill -9 <slp_pid>
2) # stopsrc -s cimsys
3) # stopsrc -s platform_agent
4) To prevent the slp_srvreg server from starting when the system is
rebooted, comment out the following two lines in /etc/inittab:
From:
platform_agent:2nce:/usr/bin/startsrc -s platform_agent >/dev/null
2>&1
cimservices:2nce:/usr/bin/startsrc -s cimsys >/dev/null 2>&1
To (use a colon : for the comment sign in the /etc/inittab file):
: platform_agent:2nce:/usr/bin/startsrc -s platform_agent >/dev/null
2>&1
: cimservices:2nce:/usr/bin/startsrc -s cimsys >/dev/null 2>&1

Action Plan: Closing Secondary & Leaving Primary with Nick for Closure.
------------
+WSAS ROBOT ID         -5765G6200  -L103/-------P2S2-13/12/23-12:43--AT

Netcom: Assigned to M Aly
+WSAS ROBOT ID         -5765G6200  -L103/-------P2S2-13/12/23-12:43--AL
+GIMBEN, NICK          -5765G6200  -L16A/-------P2S2-13/12/23-12:47--AT

Customer Rep:   Denis

Action Taken: /var/adm/messages is filling up /var.

Conf NETC,  Mihammed

Had him check if they are using Systems Director,  he says no.

So,

from PSDB:

The error you are receiving is related to Service Location Protocol
(SLP).

From the snap:
# ps -ef | grep slp_srvreg
     UID      PID     PPID   C    STIME    TTY  TIME CMD
    root  4194436        1   0 13:01:01      -  0:09 ./slp_srvreg -D

# lssrc -a
Subsystem         Group            PID          Status
platform_agent                    3670134      active
cimsys                            3211398      active

Please Note:
- SLP provides a flexible and scalable framework for providing hosts
with access to information about the existence, location, and
configuration of network services.
- The SLP service is primarily used by IBM Systems Director.

If you are not using IBM Systems Director, you can go ahead and disable
the SLP service using the following steps:
1) # slp_srvreg -k
If the slp_srvreg server is still running after using "slp_srvreg -k",
kill it manually with a kill -9:
# ps -ef | grep slp_srvreg
# kill -9 <slp_pid>
2) # stopsrc -s cimsys
3) # stopsrc -s platform_agent
4) To prevent the slp_srvreg server from starting when the system is
rebooted, comment out the following two lines in /etc/inittab:
From:
platform_agent:2nce:/usr/bin/startsrc -s platform_agent >/dev/null
2>&1
cimservices:2nce:/usr/bin/startsrc -s cimsys >/dev/null 2>&1
To (use a colon : for the comment sign in the /etc/inittab file):
: platform_agent:2nce:/usr/bin/startsrc -s platform_agent >/dev/null
2>&1
: cimservices:2nce:/usr/bin/startsrc -s cimsys >/dev/null 2>&

He followed those steps,  and the file is no longer growing.
作者: nds521    时间: 2015-01-09 11:59
coolcat982 发表于 2013-12-25 18:28
- I asked if /var/adm was filling up...  ->
it's redirecting to envision central server,it's fillin ...

时隔多年，这个帖子终于可以结了！！！感谢各位朋友热心帮助！！



欢迎光临 Chinaunix (http://bbs.chinaunix.net/) Powered by Discuz! X3.2