分类： 系统运维
2013-05-10 16:18:59
V9.7需要安装 dev-libs/libaio
|
emerge dev-libs/libaio |
V9.5需要安装 libstdc++
|
emerge virtual/libstdc++ |
|
tar -xvzf v9.5fp8_linuxx64_server.tar.gz cd server ./db2_install -b /opt/ibm/db2/V9.5 |
升级为fp10，则执行
tar –xvzf v9.5fp10_linuxx64_universal_fixpack.tar.gz
cd universal
|
./installFixPack -f db2lib -b /opt/ibm/db2/V9.5 |
|
groupadd -g 999 db2iadm1 groupadd -g 998 db2fadm1 groupadd -g 997 db2sadm1 |
|
useradd -u 1004 -g db2iadm1 -m -d /home/db2inst1 db2inst1 -p password1 useradd -u 1003 -g db2fadm1 -m -d /home/db2fenc1 db2fenc1 -p password2 useradd -u 1002 -g db2sadm1 -m -d /home/db2admn1 db2admn1 -p password3 |
注意：如果没有 "db2inst1 50000/tcp" 则会 db2start 时候报告 SQL5043N 错误，没有其他这无法启动实例
|
echo "#Add DB2 FCM port information" >> /etc/services echo "#" >> /etc/services echo "db2inst1 50000/tcp" >> /etc/services echo "DB2_db2inst1 60000/tcp" >> /etc/services echo "DB2_db2inst1_1 60001/tcp" >> /etc/services echo "DB2_db2inst1_2 60002/tcp" >> /etc/services echo "DB2_db2inst1_END 60003/tcp" >> /etc/services |
|
/opt/ibm/db2/V9.5/instance/db2icrt -p db2inst1 -u db2fenc1 db2inst1 |
如果安装实例失败，则用下列命令删除实例后重装
|
/opt/ibm/db2/V9.5/instance/db2idrop db2inst1 rm -rf /home/db2inst1/* nano -w /etc/conf.d/hostname nano -w /etc/hosts /opt/ibm/db2/V9.5/instance/db2icrt -p db2inst1 -u db2fenc1 db2inst1 |
|
/opt/ibm/db2/V9.5/instance/dascrt -u db2admn1 |
|
su - db2inst1 db2 update dbm cfg using SVCENAME db2inst1 db2set DB2COMM=tcpip db2start db2sampl |
|
db2 LIST DB DIRECTORY db2 CONNECT TO SAMPLE db2 list tables db2level |
新增许可文件
|
db2licm -a db2ese_cV9.5CPU.lic |
查看许可情况
|
db2licm -l |
|
Product name: "DB2 Enterprise Server Edition" License type: "CPU Option" Expiry date: "Permanent" Product identifier: "db2ese" Version information: "9.5" Enforcement policy: "Soft Stop" Features: DB2 Database Partitioning: "Not licensed" DB2 Performance Optimization ESE: "Not licensed" DB2 Storage Optimization: "Not licensed" DB2 Advanced Access Control: "Not licensed" DB2 Geodetic Data Management: "Not licensed" IBM Homogeneous Replication ESE: "Not licensed"
Product name: "DB2 Connect Server" License type: "Trial" Expiry date: "08/31/2010" Product identifier: "db2consv" Version information: "9.5" |
删除无用的许可
|
db2licm -r "db2consv" 到此DB2安装完成。我只做到这一步，下面步骤除去第15步，其他暂时未执行 |
禁止自动发现
|
db2 update dbm cfg using DISCOVER_INST disable db2 update dbm cfg using DISCOVER disable db2 update admin cfg using DISCOVER disable |
修改内核参数 /etc/sysctl.conf 内容如下
|
# added by caviler # for db2 kernel.shmmax = 6884901888 kernel.msgmni = 1024 kernel.sem = 250 256000 32 1024 |
使修改后的内核参数生效
|
sysctl –p |
|
db2stop exit /opt/ibm/db2/V9.5/instance/dasdrop /opt/ibm/db2/V9.5/instance/db2idrop db2inst1 rm -rf /home/db2inst1/* ./db2_deinstall -a -b /opt/ibm/db2/V9.5 |
|
nano /etc/init.d/db2 chmod 755 /etc/init.d/db2 /home/db2inst1/sqllib/bin/db2iauto -on db2inst1 nano /etc/conf.d/local.start， 在其中配置/etc/init.d/db2 start |
/etc/init.d/db2 文件内容如下
|
#!/sbin/runscript
# Path to the DB2 installation DB2DIR="/opt/ibm/db2/V9.5"
# User name of the administration instance owner DASUSR=$(${DB2DIR}/instance/daslist)
# List of database instances DB2INSTS=$(${DB2DIR}/instance/db2ilist)
depend() { need net use dns clock logger ldap }
find_homedir() { getent passwd $1 | cut -d":" -f 6 }
start() { LOGFILE=$(mktemp) for db2inst in ${DB2INSTS}; do INSTSTART=$(${DB2DIR}/instance/db2iset DB2AUTOSTART -i ${db2inst}) if [[ ${INSTSTART} == "TRUE" ]] || [[ ${INSTSTART} == "YES" ]]; then ebegin "Starting IBM DB2 database instance ${db2inst}" INSTHOME="$(find_homedir ${db2inst})" su ${db2inst} -l -c "\ source ${INSTHOME}/sqllib/db2profile ; \ db2start" >${LOGFILE} 2>&1 rc=${?} eend ${rc} [[ ${rc} -ne 0 ]] && cat ${LOGFILE} fi done if [[ -n "$DASUSR" ]]; then ebegin "Starting IBM DB2 administration instance ${DASUSR}" INSTHOME="$(find_homedir ${DASUSR})" su ${DASUSR} -l -c "\ source ${INSTHOME}/das/dasprofile ; \ db2admin start" >${LOGFILE} 2>&1 rc=${?} eend ${rc} [[ ${rc} -ne 0 ]] && cat ${LOGFILE} fi rm -f ${LOGFILE} }
stop() { LOGFILE=$(mktemp) if [[ -n "${DASUSR}" ]]; then ebegin "Stopping IBM DB2 administration instance ${DASUSR}" INSTHOME="$(find_homedir ${DASUSR})" su ${DASUSR} -c "\ source ${INSTHOME}/das/dasprofile ; \ db2admin stop" >${LOGFILE} 2>&1 rc=${?} eend ${rc} [[ ${rc} -ne 0 ]] && cat ${LOGFILE} fi for db2inst in ${DB2INSTS}; do ebegin "Stopping IBM DB2 database instances ${db2inst}" INSTHOME="$(find_homedir ${db2inst})" su - ${db2inst} -c "\ source ${INSTHOME}/sqllib/db2profile ; \ db2stop force" >${LOGFILE} 2>&1 rc=${?} eend ${rc} [[ ${rc} -ne 0 ]] && cat ${LOGFILE} done rm -f ${LOGFILE} } |
|
db2set DB2_CAPTURE_LOCKTIMEOUT=ON db2set DB2_SKIPINSERTED=ON db2set DB2_EVALUNCOMMITTED=ON db2set DB2CONNECT_IN_APP_PROCESS=NO db2set DB2COMM=tcpip db2set DB2CODEPAGE=1386 db2set DB2AUTOSTART=YES |
恢复数据库之前要做好磁盘空间规划，外网DB2是8块600G硬盘，内部测试环境则为6块1T硬盘。因为服务器使用软Raid10，所以规划时需要在外网使用df –lh，查看外网的磁盘规划是什么样的。然后在内网做出来的分区要与外网一样。大概计算方式为：
比如一个分区做完Raid10后为80G,实际上占用磁盘分区大小是160G,8块磁盘的话，也就是每个磁盘上一个分区分出20G个做Raid10，做完Raid10后为80G。
6块磁盘的话大概就是Raid10之前的160G/6块磁盘=6块磁盘上某分区的大小。
从外网DB2服务器上拷贝DB2的备份文件到测试环境，备份文件为：
/data/bak/UNA_
db2inst1@db2 ~ $ db2 RESTORE DB UNA_WWW FROM /data/bak/ TAKEN AT 20130219000003当看到successful字样，说明恢复成功（此过程非常漫长，建议放到晚上进行）
db2inst1@db2 ~ $ db2 connect to UNA_WWW
报错：SQL1117N A connection to or activation of database "UNA_WWW" cannot be made
because of ROLL-FORWARD PENDING. SQLSTATE=57019
原因：此原因是数据库恢复成功，但需要做日志回滚操作才能正常连接数据库。
解决：db2inst1@db2 ~ $ db2 rollforward db UNA_WWW to end of logs
报错：SQL4970N Roll-forward recovery on database "UNA_WWW" cannot reach the
specified stop point (end-of-log or point-in-time) on database partition(s)
"0". Roll-forward recovery processing has halted on log file "S0363523.LOG".
原因：提示恢复到end of logs需要日志文件S0363523.LOG
解决：db2inst1@db2 ~ $ db2 "restore database UNA_WWW logs from '/data/bak/' taken at 20130219000003 logtarget '/data/bak/logs/'"
如/data/bak/logs目录不存在，则新建，注意权限！当出现如下字样，说明提取日志成功：
DB20000I The RESTORE DATABASE command completed successfully.
继续进行日志回滚操作：
db2inst1@db2 ~ $ db2 "rollforward database UNA_WWW to end of logs and COMPLETE overflow log path('/data/bak/logs')"
报错：SQL1218N There are no pages currently available in bufferpool "". SQLSTATE=57011
原因：由于恢复的数据库buffer pool的设置很大，测试机器内存只有32G,正式环境内存则比这个大
解决：db2inst1@db2 ~ $ db2set DB2_OVERRIDE_BPF=500
再次进行日志回滚操作：
db2inst1@db2 ~ $ db2 "rollforward database UNA_WWW to end of logs and complete overflow log path('/data/bak/logs')"
提示如下，则表示回滚成功：
Rollforward Status
Input database alias = UNA_WWW
Number of nodes have returned status = 1
Node number = 0
Rollforward status = not pending
Next log file to be read =
Log files processed = S0363523.LOG - S0363572.LOG
Last committed transaction = 2013-02-19-07.53.47.000000 Local
DB20000I The ROLLFORWARD command completed successfully.
连接数据库：
db2inst1@db2 ~ $ db2 connect to UNA_WWW
Database Connection Information
Database server = DB2/LINUXX8664 9.5.10
SQL authorization ID = DB2INST1
Local database alias = UNA_WWW
