Rhel7_Oracle11g_rac安装方法是什么
这篇文章主要讲解了“Rhel7_Oracle11g_rac安装方法是什么”，文中的讲解内容简单清晰，易于学习与理解，下面请大家跟着小编的思路慢慢深入，一起来研究和学习“Rhel7_Oracle11g_rac安装方法是什么”吧！
1、禁用selinux
getenforcesetenforce0vim/etc/selinux/config
2、关闭防火墙、禁止开机启动
systemctlstopfirewalld.servicesystemctldisablefirewalld.service
3、修改主机名
主机名中禁止使用下划线“_”，建议使用小写字母，长度小于8位hostnamectlset-hostnamemydb1hostnamectlset-hostnamemydb2修改完重新登录
4、配置yum
mount-tiso9660-oloop/dev/sr0/media/catrhel7.repo[base]name=rhel7.7baseurl=file:///mediaenable=1gpgcheck=0
5、关闭没必要的服务
Redhat6:serviceiptablesstopserviceip6tablesstopchkconfigiptablesoffchkconfigip6tablesoffservicesshdstartchkconfigsshdonserviceBluetoothstopchkconfigBluetoothoffservicepostfixstopchkconfigpostfixoffservicecupsstopchkconfigcupsoffservicecpuspeedoffchkconfigcpuspeedoffserviceNetworkManagerstopchkconfigNetworkManageroffservicevsftpdstopchkconfigvsftpdoffservicedhcpdstopchkconfigdhcpdoffservicenfsstopchkconfignfsoffservicenfslockstopchkconfignfslockoffserviceypbindstopchkconfigypbindoffRedhat7:.................
5、安装依赖包
--检查rpm-q--qf'%{NAME}-%{VERSION}-%{RELEASE}(%{ARCH})\n'\binutils\compat-libcap1\compat-libstdc++-33\gcc\gcc-c++\glibc\glibc-devel\ksh\libstdc++\libstdc++-devel\libaio\libaio-devel\make\sysstat--安装yum-yinstallbinutilscompat-libstdc++-33gccgcc-c++glibcglibc-commonglibc-develkshlibaiolibaio-devellibgcclibstdc++libstdc++-develmakesysstatopenssh-clientscompat-libcap1xorg-x11-utilsxorg-x11-xauthelfutilsunixODBCunixODBC-devellibXpelfutils-libelfelfutils-libelf-develsmartmontools
6、内核参数修改
--计算方法：cat/proc/sys/fs/file-max+512*process*instancenumberfs.file-max=6815744kernel.sem=25032000100128kernel.shmmni=4096--计算方法：getconfPAGE_SIZETOTALRAMINBYTES/PAGE_SIZEkernel.shmall=536870912--计算方法：HALFOFTOTALRAMINBYTESkernel.shmmax=1073741824net.ipv4.ip_local_port_range=900065500net.core.rmem_default=262144net.core.wmem_default=262144net.core.rmem_max=4194304net.core.wmem_max=1048576fs.aio-max-nr=4194304vm.dirty_ratio=20vm.dirty_background_ratio=3vm.dirty_writeback_centisecs=100vm.dirty_expire_centisecs=500vm.swappiness=10vm.min_free_kbytes=524288#rp_filter，这里假设eth3和eth4都是私有网卡net.ipv4.conf.ens39.rp_filter=2net.ipv4.conf.ens39.rp_filter=2#IP分片汇聚的最大/最小内存用量，计算公式：numCPU*130000，逻辑cpu为96，那么high参数建议至少设置为12m以上。同时low参数比high参数少1m即可。net.ipv4.ipfrag_high_thresh=16777216net.ipv4.ipfrag_low_thresh=15728640net.ipv4.ipfrag_time=60-----修改后如下：fs.file-max=6815744kernel.sem=25032000100128kernel.shmmni=4096kernel.shmall=536870912kernel.shmmax=1073741824net.ipv4.ip_local_port_range=900065500net.core.rmem_default=262144net.core.wmem_default=262144net.core.rmem_max=4194304net.core.wmem_max=1048576fs.aio-max-nr=4194304vm.dirty_ratio=20vm.dirty_background_ratio=3vm.dirty_writeback_centisecs=100vm.dirty_expire_centisecs=500vm.swappiness=10vm.min_free_kbytes=524288net.ipv4.conf.ens39.rp_filter=2net.ipv4.conf.ens39.rp_filter=2net.ipv4.ipfrag_high_thresh=16777216net.ipv4.ipfrag_low_thresh=15728640net.ipv4.ipfrag_time=60
7、关闭操作系统透明大页
redhat7#cat/sys/kernel/mm/transparent_hugepage/enabledalwaysmadvise[never]--增加transparent_hugepage=nevercat/etc/default/grubGRUB_TIMEOUT=5GRUB_DISTRIBUTOR="$(sed's,release.*$,,g'/etc/system-release)"GRUB_DEFAULT=savedGRUB_DISABLE_SUBMENU=trueGRUB_TERMINAL_OUTPUT="console"GRUB_CMDLINE_LINUX="rd.lvm.lv=myvg/swaprd.lvm.lv=myvg/usrvconsole.font=latarcyrheb-sun16rd.lvm.lv=myvg/rootcrashkernel=autovconsole.keymap=usrhgbquiettransparent_hugepage=never"GRUB_DISABLE_RECOVERY="true"--对grub生效#grub2-mkconfig-o/boot/grub2/grub.cfgGeneratinggrubconfigurationfile...Foundlinuximage:/boot/vmlinuz-3.10.0-123.el7.x86_64Foundinitrdimage:/boot/initramfs-3.10.0-123.el7.x86_64.imgFoundlinuximage:/boot/vmlinuz-0-rescue-41c535c189b842eea5a8c20cbd9bff26Foundinitrdimage:/boot/initramfs-0-rescue-41c535c189b842eea5a8c20cbd9bff26.imgdone--关闭tuned服务#systemctlstoptuned.service#systemctldisabletuned.service--重启reboot--确认cat/sys/kernel/mm/transparent_hugepage/enabledredhat6、redhat7在/etc/rc.d/rc.local文件中加入下面配置，重启生效：iftest-f/sys/kernel/mm/transparent_hugepage/enabled;thenechonever>/sys/kernel/mm/transparent_hugepage/enabledfiiftest-f/sys/kernel/mm/transparent_hugepage/defrag;thenechonever>/sys/kernel/mm/transparent_hugepage/defragfi重启后确认结果如下：cat/sys/kernel/mm/transparent_hugepage/defragalways[never]cat/sys/kernel/mm/transparent_hugepage/enabledalways[never]
8、ntp设置
Redhat6配置：#vi/etc/sysconfig/ntpd#Droproottoid'ntp:ntp'bydefault.OPTIONS="-x-untp:ntp-p/var/run/ntpd.pid"#启动微调模式#Setto'yes'tosynchwclockaftersuccessfulntpdateSYNC_HWCLOCK=yes#同步硬件bios时间#vi/etc/ntp.confserverxx.xx.xx.xxprefer#添加ntp服务器地址为首要地址server127.127.1.0iburst#添加本机为次要同步地址#vi/etc/ntp/step-tickersxx.xx.xx.xx#添加ntp服务器地址，设置在ntp启动时自动同步时间配置完毕后，重启ntpd服务，并查看状态chkconfigntpdonservicentpdrestartntpstat#查看ntpd服务状态date#查看时间是否正常redhat7配置：yuminstallntpvi/etc/ntp.confserver10.5.26.10iburstvi/etc/sysconfig/ntpdOPTIONS="-x-g"systemctlstartntpd.servicesystemctlenablentpd.servicesystemctlstatusntpd.service
9、网卡绑定
red6touch/etc/modprobe.d/bonding.confecho"aliasbondeth0bonding">>/etc/modprobe.d/bonding.confvi/etc/sysconfig/network-scripts/ifcfg-bond0DEVICE=bond0ONBOOT=yesBOOTPROTO=staticUSERCTL=noNM_CONTROLLED=noIPADDR=10.1.2.3NETMASK=255.255.255.0GATEWAY=10.1.2.254BONDING_OPTS="mode=1miimon=100"vi/etc/sysconfig/network-scripts/ifcfg-eth0DEVICE=eth0ONBOOT=yesBOOTPROTO=noneMASTER=bond0SLAVE=yesUSERCTL=noNM_CONTROLLED=noecho"ifenslavebond0eth0eth2">>/etc/rc.d/rc.local重启服务器查看状态cat/sys/class/net/bonding_masterscat/sys/class/net/bondeth0/bonding/modecat/proc/net/bonding/bondeth0通过拔插网线验证red7nmcliteam
10、网络检查
(1).确保节点间通信网络（私有网络）是通过单独的交换机连接，而不是直连。(2).确保所有节点的连接到相同网络的网卡名称、网络子网都一样。比如连接到public网络的网卡名称都叫eth0，其IP地址子网都是133.37.x.0，子网掩码都为255.255.255.0。(3).确保系统中有且只有一个默认路由，并且是通过public网络到达默认路由。(4).确保网卡到网络的带宽是正确的。#ethtooleth2Settingsforeth2:Supportedports:[TP]Supportedlinkmodes:10baseT/Half10baseT/Full100baseT/Half100baseT/Full1000baseT/FullSupportsauto-negotiation:YesAdvertisedlinkmodes:10baseT/Half10baseT/Full100baseT/Half100baseT/Full1000baseT/FullAdvertisedauto-negotiation:YesSpeed:1000Mb/sDuplex:FullPort:TwistedPairPHYAD:1Transceiver:internalAuto-negotiation:onSupportsWake-on:umbgWake-on:gCurrentmessagelevel:0x00000003(3)Linkdetected:yes上面的Speed:1000Mb/s表示连接的实际网络带宽为1000Mbps。虽然交换机、网卡都是1000Mbps或以上带宽，但有时由于端口问题、网络线缆问题等原因，实际带宽并没有这么多。(5).确保在私有网络上开启了多播。可以在Oracle官方支持网站上下载mcasttest.pl脚本进行检查
11、存储多路径配置
见https://www.modb.pro/db/14031
12、创建用户组
groupadd-g1000oinstallgroupadd-g1001dbagroupadd-g1002opergroupadd-g1003asmadmingroupadd-g1004asmopergroupadd-g1005asmdbauseradd-u1000-goinstall-Gdba,oper,asmdbaoracleuseradd-u1001-goinstall-Gdba,asmadmin,asmdba,asmoper,opergrid
13、limits限制
touch/etc/security/limits.d/99-grid-oracle-limits.confgridsoftnproc16384gridhardnproc16384gridsoftnofile10240gridhardnofile65536gridsoftstack10240gridhardstack32768gridsoftmemlockunlimitedgridhardmemlockunlimitedgridsoftcoreunlimitedgridhardcoreunlimitedoraclesoftnproc16384oraclehardnproc16384oraclesoftnofile10240oraclehardnofile65536oraclesoftstack10240oraclehardstack32768oraclesoftmemlockunlimitedoraclehardmemlockunlimitedoraclesoftcoreunlimitedoraclehardcoreunlimitedtouch/etc/profile.d/oracle-grid.sh#Settingtheappropriateulimitsfororacleandgriduserif[$USER="oracle"];thenif[$SHELL="/bin/ksh"];thenulimit-u16384ulimit-n65536elseulimit-u16384-n65536fifiif[$USER="grid"];thenif[$SHELL="/bin/ksh"];thenulimit-u16384ulimit-n65536elseulimit-u16384-n65536fifi
14、创建目录
集群软件BASE目录：/u01/app/oracle集群软件HOME目录：/u01/app/11.2.0/grid数据库软件BASE目录：/u01/app/oracle数据库软件HOME目录：/u01/app/oracle/product/11.2.0/db_homemkdir-p/u01/app/gridmkdir-p/u01/app/11.2.0/gridmkdir-p/u01/app/oraInventorychown-Rgrid:oinstall/u01/mkdir-p/u01/app/oracle/product/11.2.0/db_homechown-Roracle:oinstall/u01/app/oracle/chmod-R755/u01
15、设置oracle、grid用户环境变量
oracle:exportTMP=/tmpexportTMPDIR=$TMPexportORACLE_SID=racdb1;exportORACLE_BASE=/u01/app/oracleexportORACLE_HOME=$ORACLE_BASE/product/11.2.0/db_homeexportORACLE_TERM=xtermexportPATH=/usr/sbin:$PATHexportPATH=$ORACLE_HOME/bin:$PATHexportLD_LIBRARY_PATH=$ORACLE_HOME/lib:/lib:/usr/libexportCLASSPATH=$ORACLE_HOME/JRE:$ORACLE_HOME/jlib:$ORACLE_HOME/rdbms/jlibexportNLS_LANG=AMERICAN_AMERICA.ZHS16GBKexportLANG=en_US.UTF-8umask022grid:exportORACLE_SID=+ASM1exportORACLE_OWNER=gridexportORACLE_BASE=/u01/app/gridexportORACLE_HOME=/u01/app/11.2.0/gridexportLD_LIBRARY_PATH=$LD_LIBRARY_PATH:$ORACLE_HOME/lib:/lib:/usr/lib:/usr/local/libPATH=$PATH:$HOME/bin$:$ORACLE_HOME/binexportLANG=en_US.UTF-8exportPATHumask022
16、修改hosts文件
PublicIP：别名直接使用主机名，即uname–a返回的机器名；PrivateIP:别名为主机名-priv。VirtualIP：别名为主机名-vip。SCANIP：别名为数据库名-scan。#PublicIp192.168.0.203mydb1192.168.0.204mydb2#VirtualIp192.168.0.205mydb1-vip192.168.0.206mydb2-vip#PrivateIp192.168.124.203mydb1-priv192.168.124.204mydb2-priv#ScanIp192.168.0.207racdb-scan
17、设置互信
安装grid软件时使用sshUserSetup.sh快速创建互信,$node1$node2变量参数需要根据实际节点名称进行内容调整。在一个节点上执行即可（可以在root用户下执行）：./sshUserSetup.sh-usergrid-hosts"$node1$node2"-advanced-exverify–confirm./sshUserSetup.sh-useroracle-hosts"$node1$node2"-advanced-exverify-confirm
18、命名规范
1）集群（CLUSTER）的命名规则Clustername本身没有特殊用途，仅在使用其他管理工具统一管理不同RAC实例时有用，不得超过15个字符。${DB_NAME}-cls2)SCAN的命名规则SCAN名称本身没有特殊用途，仅在使用其他管理工具统一管理时有用，不得超过15个字符。${DB_NAME}-scan
19、安装时检查忽略选项
1）Package：pdksh-5.2.142)DeviceChecksforAM3)Taskresolv.confIntegrity
20、grid集群软件安装，在执行root.sh脚本前，添加ohas服务
touch/usr/lib/systemd/system/ohas.servicechmod777/usr/lib/systemd/system/ohas.serviceohas.service：[Unit]Description=OracleHighAvailabilityServicesAfter=syslog.target[Service]ExecStart=/etc/init.d/init.ohasdrun>/dev/null2>&1Type=simpleRestart=always[Install]WantedBy=multi-user.targetsystemctldaemon-reloadsystemctlenableohas.servicesystemctlstartohas.servicesystemctlstatusohas.service
21、在两个节点分别执行orainstRoot.sh和root.sh脚本
执行顺序：1节点orainstRoot.sh，2节点orainstRoot.sh，1节点root.sh,2节点root.sh1节点：[root@mydb1system]#/u01/app/oraInventory/orainstRoot.shChangingpermissionsof/u01/app/oraInventory.Addingread,writepermissionsforgroup.Removingread,write,executepermissionsforworld.Changinggroupnameof/u01/app/oraInventorytooinstall.Theexecutionofthescriptiscomplete.[root@mydb1system]#/u01/app/11.2.0/grid/root.shPerformingrootuseroperationforOracle11gThefollowingenvironmentvariablesaresetas:ORACLE_OWNER=gridORACLE_HOME=/u01/app/11.2.0/gridEnterthefullpathnameofthelocalbindirectory:[/usr/local/bin]:Copyingdbhometo/usr/local/bin...Copyingoraenvto/usr/local/bin...Copyingcoraenvto/usr/local/bin...Creating/etc/oratabfile...Entrieswillbeaddedtothe/etc/oratabfileasneededbyDatabaseConfigurationAssistantwhenadatabaseiscreatedFinishedrunninggenericpartofrootscript.Nowproduct-specificrootactionswillbeperformed.Usingconfigurationparameterfile:/u01/app/11.2.0/grid/crs/install/crsconfig_paramsCreatingtracedirectoryUserignoredPrerequisitesduringinstallationInstallingTraceFileAnalyzerOLRinitialization-successfulrootwalletrootwalletcertrootcertexportpeerwalletprofilereaderwalletpawalletpeerwalletkeyspawalletkeyspeercertrequestpacertrequestpeercertpacertpeerrootcertTPprofilereaderrootcertTPparootcertTPpeerpacertTPpapeercertTPprofilereaderpacertTPprofilereaderpeercertTPpeerusercertpausercertAddingClusterwareentriestoinittab**ohasdfailedtostart**##出现该提示需要重启ohas.service服务FailedtostarttheClusterware.Last20linesofthealertlogfollow:2020-01-1008:16:18.019:[client(23496)]CRS-2101:TheOLRwasformattedusingversion3.CRS-2672:Attemptingtostart'ora.mdnsd'on'mydb1'CRS-2676:Startof'ora.mdnsd'on'mydb1'succeededCRS-2672:Attemptingtostart'ora.gpnpd'on'mydb1'CRS-2676:Startof'ora.gpnpd'on'mydb1'succeededCRS-2672:Attemptingtostart'ora.cssdmonitor'on'mydb1'CRS-2672:Attemptingtostart'ora.gipcd'on'mydb1'CRS-2676:Startof'ora.cssdmonitor'on'mydb1'succeededCRS-2676:Startof'ora.gipcd'on'mydb1'succeededCRS-2672:Attemptingtostart'ora.cssd'on'mydb1'CRS-2672:Attemptingtostart'ora.diskmon'on'mydb1'CRS-2676:Startof'ora.diskmon'on'mydb1'succeededCRS-2676:Startof'ora.cssd'on'mydb1'succeeded已成功创建并启动ASM。已成功创建磁盘组OCRDG。clscfg:-installmodespecifiedSuccessfullyaccumulatednecessaryOCRkeys.CreatingOCRkeysforuser'root',privgrp'root'..Operationsuccessful.CRS-4256:UpdatingtheprofileSuccessfuladditionofvotingdisk992a298111ba4fb8bf16c75cdd232ca8.Successfullyreplacedvotingdiskgroupwith+OCRDG.CRS-4256:UpdatingtheprofileCRS-4266:Votingfile(s)successfullyreplaced##STATEFileUniversalIdFileNameDiskgroup------------------------------------------1.ONLINE992a298111ba4fb8bf16c75cdd232ca8(/dev/mapper/asm_ocr1p1)[OCRDG]Located1votingdisk(s).sh:/bin/netstat:没有那个文件或目录CRS-2672:Attemptingtostart'ora.asm'on'mydb1'CRS-2676:Startof'ora.asm'on'mydb1'succeededCRS-2672:Attemptingtostart'ora.OCRDG.dg'on'mydb1'CRS-2676:Startof'ora.OCRDG.dg'on'mydb1'succeeded软件包准备中...cvuqdisk-1.0.9-1.x86_64ConfigureOracleGridInfrastructureforaCluster...succeeded2节点：[root@mydb2etc]#/u01/app/oraInventory/orainstRoot.shChangingpermissionsof/u01/app/oraInventory.Addingread,writepermissionsforgroup.Removingread,write,executepermissionsforworld.Changinggroupnameof/u01/app/oraInventorytooinstall.Theexecutionofthescriptiscomplete.[root@mydb2etc]#[root@mydb2etc]#cd[root@mydb2~]#/u01/app/11.2.0/grid/root.shPerformingrootuseroperationforOracle11gThefollowingenvironmentvariablesaresetas:ORACLE_OWNER=gridORACLE_HOME=/u01/app/11.2.0/gridEnterthefullpathnameofthelocalbindirectory:[/usr/local/bin]:Copyingdbhometo/usr/local/bin...Copyingoraenvto/usr/local/bin...Copyingcoraenvto/usr/local/bin...Creating/etc/oratabfile...Entrieswillbeaddedtothe/etc/oratabfileasneededbyDatabaseConfigurationAssistantwhenadatabaseiscreatedFinishedrunninggenericpartofrootscript.Nowproduct-specificrootactionswillbeperformed.Usingconfigurationparameterfile:/u01/app/11.2.0/grid/crs/install/crsconfig_paramsCreatingtracedirectoryUserignoredPrerequisitesduringinstallationInstallingTraceFileAnalyzerOLRinitialization-successfulAddingClusterwareentriestoinittab**ohasdfailedtostart**##出现该提示需要重启ohas.service服务FailedtostarttheClusterware.Last20linesofthealertlogfollow:2020-01-1008:27:01.458:[client(21808)]CRS-2101:TheOLRwasformattedusingversion3.CRS-4402:TheCSSdaemonwasstartedinexclusivemodebutfoundanactiveCSSdaemononnodemydb1,number1,andisterminatingAnactiveclusterwasfoundduringexclusivestartup,restartingtojointheclustersh:/bin/netstat:没有那个文件或目录软件包准备中...cvuqdisk-1.0.9-1.x86_64ConfigureOracleGridInfrastructureforaCluster...succeeded
22、报错忽略
1）ConfigureOracleGridInfrastructureforaCluster2)OracleClusterVerificationUtility
23、数据库软件安装检查忽略项
1）Package:pdksh-5.2.142)Taskresolv.confIntegrity3)SingleClientAccessName(SCAN)
23、安装报错
提示：Errorininvokingtarget'agentnmhs'ofmakefile'/u01/app/oracle/product/11.2.0/db_home/sysman/lib/ins_emagent.mk'.解决方法如下：cd$ORACLE_HOME/sysman/libcpins_emagent.mkins_emagent.mk.bakviins_emagent.mk/NMECTL快速定位，修改如下：$(MK_EMAGENT_NMECTL)-lnnz11说明：第一个是字母l后面两个是数字1然后点击Retry
24、root用户下，在两个节点分别执行root.sh
/u01/app/oracle/product/11.2.0/db_home/root.sh[root@mydb2~]#/u01/app/oracle/product/11.2.0/db_home/root.shPerformingrootuseroperationforOracle11gThefollowingenvironmentvariablesaresetas:ORACLE_OWNER=oracleORACLE_HOME=/u01/app/oracle/product/11.2.0/db_homeEnterthefullpathnameofthelocalbindirectory:[/usr/local/bin]:Thecontentsof"dbhome"havenotchanged.Noneedtooverwrite.Thecontentsof"oraenv"havenotchanged.Noneedtooverwrite.Thecontentsof"coraenv"havenotchanged.Noneedtooverwrite.Entrieswillbeaddedtothe/etc/oratabfileasneededbyDatabaseConfigurationAssistantwhenadatabaseiscreatedFinishedrunninggenericpartofrootscript.Nowproduct-specificrootactionswillbeperformed.Finishedproduct-specificrootactions.
25、使用asmca配置实例安装所需磁盘
REDODG
DATADG
ARCHDG
26、使用dbca创建实例
27、配置hugepage，大页内存
使用大内存页有哪些好处：
1.减少页表(PageTable）大小。每一个HugePage，对应的是连续的2MB物理内存，这样12GB的物理内存只需要48KB的PageTable，与原来的24MB相比减少很多。2.HugePage内存只能锁定在物理内存中，不能被交换到交换区。这样避免了交换引起的性能影响。3.由于页表数量的减少，使得CPU中的TLB（可理解为CPU对页表的CACHE）的命中率大大提高。4.针对HugePage的页表，在各进程之间可以共享，也降低了PageTable的大小。实际上这里可以反映出Linux在分页处理机制上的缺陷。而其他操作系统，比如AIX，对于共享内存段这样的内存，进程共享相同的页表，避免了Linux的这种问题。像笔者维护的一套系统，连接数平常都是5000以上，实例的SGA在60GB左右，要是按Linux的分页处理方式，系统中大部分内存都会被页表给用掉。
实施步骤如下：
1）检查/proc/meminfogrep-ihugepage/proc/meminfo2)计算HugePages_Total大小，使用hugepages_settings.sh脚本进行计算：#!/bin/bashKERN=`uname-r|awk-F.'{printf("%d.%d\n",$1,$2);}'`#FindouttheHugePagesizeHPG_SZ=`grepHugepagesize/proc/meminfo|awk'{print$2}'`#Startfrom1pagestobeonthesafesideandguarantee1freeHugePageNUM_PG=1#CumulativenumberofpagesrequiredtohandletherunningsharedmemorysegmentsforSEG_BYTESin`ipcs-m|awk'{print$5}'|grep"[0-9][0-9]*"`doMIN_PG=`echo"$SEG_BYTES/($HPG_SZ*1024)"|bc-q`if[$MIN_PG-gt0];thenNUM_PG=`echo"$NUM_PG+$MIN_PG+1"|bc-q`fidone#Finishwithresultscase$KERNin'2.4')HUGETLB_POOL=`echo"$NUM_PG*$HPG_SZ/1024"|bc-q`;echo"Recommendedsetting:vm.hugetlb_pool=$HUGETLB_POOL";;'2.6')MEM_LOCK=`echo"$NUM_PG*$HPG_SZ"|bc-q`;echo"Recommendedsettingwithinthekernelbootcommandline(/etc/sysctl.conf):vm.nr_hugepages=$NUM_PG"echo"Recommendedsettingwithin/etc/security/limits.d/99-grid-oracle-limits.conf:oraclesoftmemlock$MEM_LOCK"echo"Recommendedsettingwithin/etc/security/limits.d/99-grid-oracle-limits.conf:oraclehardmemlock$MEM_LOCK";;'3.10')MEM_LOCK=`echo"$NUM_PG*$HPG_SZ"|bc-q`;echo"Recommendedsettingwithinthekernelbootcommandline(/etc/sysctl.conf):vm.nr_hugepages=$NUM_PG"echo"Recommendedsettingwithin/etc/security/limits.d/99-grid-oracle-limits.conf:oraclesoftmemlock$MEM_LOCK"echo"Recommendedsettingwithin/etc/security/limits.d/99-grid-oracle-limits.conf:oraclehardmemlock$MEM_LOCK";;*)echo"Unrecognizedkernelversion$KERN.Exiting.";;esac#end----3）修改/etc/sysctl.conf文件，增加如下行，根据上步计算的hugepages大小：vm.nr_hugepages=92184)生效sysctl-p5）修改/etc/security/limits.d/99-grid-oracle-limits.conf，增加如下,设定oracle用户可以锁定内存的大小，以KB为单位，可以设置为具体值，也可设置为unlimited:oraclesoftmemlockunlimitedoraclehardmemlockunlimited6）重新启动实例
28、修改本地和集群监听端口号为11521
1）修改前确认：[grid@mydb1~]$srvctlconfiglistenerName:LISTENERNetwork:1,Owner:gridHome:<CRShome>Endpoints:TCP:1521[grid@mydb1~]$srvctlconfigscan_listenerSCANListenerLISTENER_SCAN1exists.Port:TCP:15212）在集群下修改listener、scan_listener端口为11521,在一个节点执行srvctlmodifylistener-lLISTENER-p"TCP:11521"srvctlmodifyscan_listener-p115213）修改local_listener，进入sqlplus;altersystemsetlocal_listener='(ADDRESS=(PROTOCOL=TCP)(HOST=192.168.0.205)(PORT=11521))'scope=bothsid='racdb1';altersystemsetlocal_listener='(ADDRESS=(PROTOCOL=TCP)(HOST=192.168.0.206)(PORT=11521))'scope=bothsid='racdb2';4)修改remote_listeneraltersystemsetremote_listener='racdb-scan:11521'scope=both;5)在第1个节点关闭本地监听，修改listener.ora、endpoints_listener.ora、tnsnames.ora，并重启本地监听srvctlstoplistener-lLISTENER-nmydb1cd$ORACLE_HOME/network/admin--修改1521为11521viendpoints_listener.oravilistener.orasrvctlstartlistener-lLISTENER-nmydb1srvctlstatuslistener-lLISTENERsrvctlconfiglistener6)在第2个节点关闭本地监听，修改listener.ora、endpoints_listener.ora、tnsnames.ora，并重启本地监听srvctlstoplistener-lLISTENER-nmydb2cd$ORACLE_HOME/network/admin--修改1521为11521viendpoints_listener.oravilistener.orasrvctlstartlistener-lLISTENER-nmydb2srvctlstatuslistener-lLISTENERsrvctlconfiglistener7)修改ASM监听端口(如果不修改asm监听端口，lsnrctlstatus查看监听状态时不会显示asm服务监听状态）su-gridsqlplus/assysdbaaltersystemsetlocal_listener='(ADDRESS=(PROTOCOL=TCP)(HOST=192.168.0.205)(PORT=11521))'scope=bothsid='+ASM1';altersystemsetlocal_listener='(ADDRESS=(PROTOCOL=TCP)(HOST=192.168.0.206)(PORT=11521))'scope=bothsid='+ASM2';lsnrctlstatus
29、asm参数优化
ASM磁盘组使用的是默认的1MAU大小，对于大型数据库，这会造成较多的内存占用，同时对性能略微有些影响，建议对于新增的用于放置数据文件的ASM磁盘组，适当调大AU大小，比如4M或8M（2的幂值）。根据电信运营商的实际经验，建议设置AU为为4m。
30、数据库参数修改推荐
Altersystemsetresource_manager_plan='FORCE:'scope=spfilesid='*';Altersystemsetaudit_trail=nonescope=spfilesid='*';altersystemsetundo_retention=10800scope=spfilesid='*';altersystemsetsession_cached_cursors=200scope=spfilesid='*';altersystemsetdb_files=2000scope=spfilesid='*';altersystemsetmax_shared_servers=0scope=spfilesid='*';altersystemsetsec_max_failed_login_attempts=100scope=spfilesid='*';altersystemsetdeferred_segment_creation=falsescope=spfilesid='*';altersystemsetparallel_force_local=truescope=spfilesid='*';altersystemsetparallel_max_servers=32scope=spfilesid='*';altersystemsetsec_case_sensitive_logon=falsescope=spfilesid='*';altersystemsetopen_cursors=3000scope=spfilesid='*';altersystemsetopen_links=40scope=spfilesid='*';altersystemsetopen_links_per_instance=40scope=spfilesid='*';altersystemsetdb_cache_advice=offscope=spfilesid='*';altersystemset"_b_tree_bitmap_plans"=falsescope=spfilesid='*';altersystemset"_gc_policy_time"=0scope=spfilesid='*';altersystemset"_gc_defer_time"=3scope=spfilesid='*';altersystemset"_lm_tickets"=5000scope=spfilesid='*';altersystemset"_optimizer_use_feedback"=falsesid='*';altersystemset"_undo_autotune"=falsescope=bothsid='*';altersystemset"_bloom_filter_enabled"=FALSEscope=spfilesid='*';altersystemset"_cleanup_rollback_entries"=2000scope=spfilesid='*';altersystemset"_px_use_large_pool"=truescope=spfilesid='*';altersystemset"_optimizer_extended_cursor_sharing_rel"=NONEscope=spfilesid='*';altersystemset"_optimizer_extended_cursor_sharing"=NONEscope=spfilesid='*';altersystemset"_optimizer_adaptive_cursor_sharing"=falsescope=spfilesid='*';altersystemset"_optimizer_mjc_enabled"=FALSEscope=spfilesid='*';altersystemset"_sort_elimination_cost_ratio"=1scope=spfilesid='*';altersystemset"_partition_large_extents"=FALSEscope=spfilesid='*';altersystemset"_index_partition_large_extents"=FALSEscope=spfilesid='*';altersystemset"_clusterwide_global_transactions"=FALSEscope=spfilesid='*';altersystemset"_part_access_version_by_number"=FALSEscope=spfile;altersystemset"_partition_large_extents"=FALSEscope=spfile;altersystemset"_sort_elimination_cost_ratio"=1scope=spfile;altersystemset"_use_adaptive_log_file_sync"=FALSEscope=spfile;altersystemset"_lm_sync_timeout"=1200scope=spfile;altersystemset"_ksmg_granule_size"=134217728scope=spfile;altersystemset"_optimizer_cartesian_enabled"=falsescope=spfile;altersystemset"_external_scn_logging_threshold_seconds"=3600scope=spfile;altersystemset"_datafile_write_errors_crash_instance"=falsescope=spfile;altersystemsetevent='28401TRACENAMECONTEXTFOREVER,LEVEL1:60025tracenamecontextforever:10949tracenamecontextforever,level1'sid='*'scope=spfile;
感谢各位的阅读，以上就是“Rhel7_Oracle11g_rac安装方法是什么”的内容了，经过本文的学习后，相信大家对Rhel7_Oracle11g_rac安装方法是什么这一问题有了更深刻的体会，具体使用情况还需要大家实践验证。这里是亿速云，小编将为大家推送更多相关知识点的文章，欢迎关注！
声明：本站所有文章资源内容，如无特殊说明或标注，均为采集网络资源。如若本站内容侵犯了原著者的合法权益，可联系本站删除。
