检查数据库连接性

 

 

输入：数据库名字
输出：连接数据库测试信息


#!/bin/ksh
# 
# Description : Check Status of Oracle Instance (All databases | a specific one )  , Listener, Oracle Net connection
# 
# Date of Modification         Modification
# ===========================  ===============================
# 12-AUG-2010   Sing                                   
#  
#set -x
PATH=/usr/local/bin:/usr/bin:$PATH:.

. $HOME/utility/macro/macro.env

OVO_DIR=/tmp

LOGFILE=$REPORT_DIR/chk_conn_subsystems.log_`date '+%y%m%d'`
MAILLOG=$TMP_DIR/chk_conn_subsystems.log_`date '+%y%m%d%H%M'`
ERRLOG=$TMP_DIR/chk_conn_subsystems.err
tmpfile=$TMP_DIR/chk_conn_subsystems.tmp1

OVOLOG=$OVO_DIR/chk_conn_subsystems.log
OVOSMY=$OVO_DIR/chk_conn_subsystems_smy.log

db_list="A B C D"


err_dblst=""
sname=yidbmon

EMAIL_LST=$ADMIN_DIR/email.lst

mail_flag=N
WAIT_TIME=$1

rm $OVOLOG

touch $tmpfile $OVOLOG
for db_name in ${db_list}
do
LOGIN_ID=oper/oper123
if [ "${db_name}" = "a" ] || [ "${db_name}" = "b" ]  ; then
        LOGIN_ID=oper/oper321
else if [ "${db_name}" = "c" ] || [ "${db_name}" = "d" ] ; then
         LOGIN_ID=sysman/oracle123
     fi
fi
                $ORACLE_HOME/bin/sqlplus -s ${LOGIN_ID}@${db_name} @$SCRIPT_DIR/chk_conn.sql > $tmpfile.$db_name 2>&1 & 
                sleep ${WAIT_TIME} 

                Msg=""
                dt1=`date '+%y/%m/%d %H:%M:%S'`
                dt2=`date '+%y%m%d%H%M%S'`

                grep -i $db_name"-" $tmpfile.$db_name >/dev/null

                if [ $? -ne 0 ]; then
                        mail_flag="Y"

                        echo "Unable to connect to database: " $db_name ".."${dt1} >> $MAILLOG

                        err_dblst=${err_dblst}" "$db_name"("${dt2}") " 

                else
                        echo "Connect to database successfully: " $db_name ".."${dt1} >> $MAILLOG
                fi
done

                echo "\n" >> $MAILLOG

grep -i "Unable to connect" $MAILLOG >> /dev/null


        if [ $? -eq 0 ]; then

         echo "Possible Causes:" > $ERRLOG
         echo "1. Scheduled shutdown." >> $ERRLOG 
         echo "2. Maybe the connection needs more than "${WAIT_TIME}" seconds to complete." >> $ERRLOG
         echo "3. Database or listener is not up and running." >> $ERRLOG
         echo "Actions:" >> $ERRLOG
         echo "1. Please ignore this message if it is a scheduled shutdown." >> $ERRLOG
         echo "2. Otherwise run program /macro/chk_db.sh to check database services on server side." >> $ERRLOG
         echo "\n" >> $ERRLOG

#              for iname in "dba@$EMAIL_LST"
               for iname in `cat $EMAIL_LST`
                do
                  cat $MAILLOG | grep "Unable to connect to database:" >> $ERRLOG
                  /bin/mailx -s "Failed to connect to database " $iname < $ERRLOG
                done
        else
                        rm -f $tmpfile
        fi

if [ "${err_dblst}" != "" ]; then
        echo "critical Unable to connect to db : "${err_dblst}" Check Email or log "${OVOLOG}" on server "$sname > $OVOLOG
fi

cat $MAILLOG >> $LOGFILE
cat $OVOLOG >> $OVOSMY 

chmod 777 $OVOLOG

rm -f $MAILLOG $tmpfile
find $REPORT_DIR -name chk_conn_subsystems.log_\* -mtime +5 -exec rm {} \; 

 

 

可能原因：

1. rman 异常中断。导致数据库session里的rman 程序异常。无法正常清理，强制kill session 也没用。

the issue was due to the 3rd party SBT library rman backup from the investigation, which the rman sessions in db doesn't cleared after backup

fixed:

重启备份软件，或者重启主机

 

 

2.vm 底部的杀毒软件导致连接过慢。()

fixed:

联系OA 查询

 

3.内存消耗异常

查看os log 会有告警信息出来。

fixed:

解决内存问题，或者升级oracle

 

3.db 连接偶发性连接过慢，原因未明。

 

4.监控脚本本身异常 导致系统 误报。

大量未删除的文件 导致 脚本无法连接数据库，写入日志，出现如下提示。

 

[root@yidbmon1 ~]# /usr/sbin/lsof |grep home

bash       6623    oracle  cwd       DIR              253,8      12288     131223 /home/oracle/utility/macro

tnslsnr    6940    oracle  cwd       DIR              253,8       4096     131073 /home/oracle

bash       9990    oracle  cwd       DIR              253,8       4096     131099 /home/oracle/11

bash       9990    oracle  254r      REG              253,8       3493     143496 /home/oracle/11/osw_sort_11pr.sh

bash       9992    oracle  cwd       DIR              253,8       4096     131099 /home/oracle/11

tee        9993    oracle  cwd       DIR              253,8       4096     131099 /home/oracle/11

sh        11637    oracle  cwd       DIR              253,8       4096     131099 /home/oracle/11

sh        11637    oracle    1w      REG              253,8 1994113024     147308 /home/oracle/11/nohup.out (deleted)

sh        11637    oracle    2w      REG              253,8 1994113024     147308 /home/oracle/11/nohup.out (deleted)

sh        11637    oracle  255r      REG              253,8       3493     143496 /home/oracle/11/osw_sort_11pr.sh

sleep     29934    oracle    2w      REG              253,8 1994113024     147308 /home/oracle/11/nohup.out (deleted)

sleep     29957    oracle  cwd       DIR              253,8       4096     131099 /home/oracle/11

sleep     29988    oracle  cwd       DIR              253,8       4096     131099 /home/oracle/1

 

 

6.监听程序异常，原因未明。

fixed:

以管理员身份重启监听。