Phenomenon of the problem
There are SC abnormal problems in the environment of the independent SC V6R6C10SPC800
version on site, which can't provide services normally.
2.problem analysis
Check the situation of SC service ports ** on the spot, and find that all ports are not normal **
Continue to check the system and find that the mount disk space where the SC database path /DB_DATA is
located has been occupied, causing the SC database to fail to start normally, resulting in an exception
Database disk space is occupied, the reason is to store the log logrecord tablespace large, specific view of the
storage of the log content, there are roughly the following log information
The network card does not plug the log of the network cable
An exception log for data was not received by the pipe
For the log record of the network card without inserting the network cable, the configuration error
of the system network card after on-site troubleshooting led to the actual check of the network cable
but the corresponding network card error. After adjusting the system, the log record problem was solved
For pipes that do not receive data, by looking at the log, at 23:50 SC has a log of inter-process pipe exceptions
For the current network environment, the system network card configuration is modified to solve the problem
of recording a large number of logs. For problem 2, the root causes of pipeline anomalies between software processes
caused by 23:50 need to be continuously investigated.
This problem has appeared on the SUSE system provided by IT for many times. Since there was no result of
IT inspection before and IT was a self-use environment for testing before, IT has not been effectively dealt with.
By comparing the euler system of the intelligent meeting of the company bureau with the program running on
the SUSE system provided by the company IT, the following five suspicious points are listed
1, / usr/local/HWSM /
2, / usr/local/sacc /
3, / usr/local/zmonitor/zabbix - 3.0.1 / agent/sbin/zabbix_agentd
4. Osadmin 11170 0.0 0.0 15444 1380? S 15:14 0:01 /home/osadmin/nmon/
nmon_x86_64_sles11-ft-s 60-c 526-m /home/osadmin/nmonlog
Routine. Sh's routine
1 and 2 are also present on the euler system of the corporate intelligence meeting, so the possibility of impact is ruled out
3. After consulting the maintenance personnel of the relevant detection system, the possibility of fixed 23:50
affecting the system was ruled out, and the program was run by the system users with their own low permissions,
which was less likely to affect the operation of the software
4. No corresponding personnel have been found for investigation
Check the corresponding timing task of 5, and find that the program runs with root permission, and there is a timing
task that performs operations at a fixed time of 23:45:00
Subsequently, network problems can be reproduced by manually calling the routine. Sh related operations, and the
root cause of 23:50 affecting the normal business of the software can be determined because of the scheduled tasks
on the system routine.
3. Handling Suggestions
The routine. Sh script is a tool used for performance detection, which has been abandoned. Zabbix is currently
installed on the current network for detection, so the script has no practical effect, so the current network needs to
stop the script and timed task. Now the network needs to check the four main and standby sets of teleconference
and 8 SCS, and stop routine. Sh's regular tasks
