PowerHA7.1安裝配置文檔

1、文檔記錄版本記錄安裝配置文檔Vers i on作者:Date of this revision: 2014/5/3 Saturday版本號時間變更紀(jì)要是否變更2014/5/3 Saturday形成文檔N安裝配置文檔11 規(guī)劃42 安裝軟件7拓?fù)鋱D7SAN設(shè)置（可選）7安裝軟件83 設(shè)置網(wǎng)絡(luò)相關(guān)8網(wǎng)卡IP設(shè)置8HOSTS文件設(shè)置8酉己置Vetc/cluster/rhosts 文件9準(zhǔn)備REPOSITORY DISK和共享存儲94 創(chuàng)建群集10初始化群集10定義 Repository Disk10配置 PERSISTENT IP11同步11查看配置信息12配置資源和應(yīng)用程序12Applicatio

2、ns 12配矍服務(wù)/P.13創(chuàng)速資源組14修改資源組屬性14添加卷組15再次同步165 HACMP狀態(tài)管理16啟停 HACMP16查看HACMP最終狀態(tài)17修復(fù)查看HACMP狀態(tài)不正常17刪除 REPOSITORY DISK181規(guī)劃項目要求硬件 支持PowerHA Hardware Support Matr i x:系統(tǒng)版本 要求6100-06 or 7100 補丁版本推薦軟件 要求The fol lowing Base Operating System (BOS) components for AIX are requi red for PowerHA：(requi red for the

3、 1BM Systems Di rector plug-in)The fol lowing file sets on the AIX base med i a are requi red:The appropr iate versions of RSCT for the supported AIX releases are a 1 so supplied with thePowerHA instal1 at ion med i a.硬件 要求Have at least two network adapters connected to different Ethernet switches i

4、n the same vi rtua1 LAN (VLAN)Emp1oy dual-fabric SAN connections to the storage subsystems using at least two Fibre Channe 1 (FC) adapters and appropr iate multipath dr i vers.網(wǎng)絡(luò) 要求3個網(wǎng)段，不再支持IP替換的方式共享 存儲支持主流SAN存儲PowerHA supports on 1 y the use of enhanced concurrent vo 1 ume groupsSAN心跳交換機上面要劃zone, H

5、BA卡要設(shè)置（查看repositor y disk存儲上的磁盤，大小512MW460G。不支持LVM鏡像。 后支持第三方多路徑資料P0WREHA信息中心：相關(guān)資源下莪（powerha資祝匯總）：紅皮書：窗IBM PowerHA SystemMirror 7.1 for AIX.pdfIBM PowerHA SystemMirror Standard Edition 7.1.1 for AIX Update.pdf2安裝軟件2.1拓?fù)鋱D2.2 SAN設(shè)置（可選）zone i ng:兩個SAN交換機上分別創(chuàng)建主機到主機的zoneo修改HBA卡屬性：方法一：適用于可以刪除子設(shè)備時rmdev -RI

6、fcsXchdev -I fcsX -a tme=yeschdev -I fscsiX -a dyntrk=yes -a fc_err_recov=fast_faiIcfgmgr -I fcsX；cfgmgr -I sfwco<rmX臉證：Isdev -C | grep -e fcsX -e sfwcomnXIsattr -El fcsX | grep tmeIsattr -El fscsiX | grep -e dyntrk -e fc_err_recov方法二：適用于設(shè)備忙時，重啟生效chdev -P -I fcsX -a tme=yeschdev -P -I fscsiX -a d

7、yntrk=yes -a fc_err_recov=fast_faiI2.3安裝軟件最簡安裝包（待測試）3設(shè)置網(wǎng)絡(luò)相關(guān)3.1網(wǎng)卡IP設(shè)置IP標(biāo)簽地址網(wǎng)絡(luò)接口網(wǎng)絡(luò)名接口功能IP地址網(wǎng)絡(luò)掩碼Syndey_enOenOetherl#boot1Syndey _en1en1etherl#boot2Syndeyetherl#PersistentSyndey _svcetherl#servicePerth_enOenOetherl#boot1Perth_en1en1etherl#boot2Perthetherl#PersistentPerth_svcetherlftservice3. 2 hosts文件設(shè)置

8、主機名和節(jié)點名對應(yīng)，/etc/hosts文件中必須要有主機名對應(yīng)的IP 每臺服務(wù)器都要修改Syndey_enO#boot1Syndey_en1#boot2Syndey#PersistentSyndey_svc#servicePerth_enO#boot1Perth_en1#boot2Perth#PersistentPerth svc#service由于要求配置cluster前主機名能解析。通過下面方法確認(rèn):# hostnameIxargs ping <<正常返回就,可以PING Syndey： 0： 56 data bytes64 bytes from : icmp_seq=0 t

9、tl=255 time=O ms如果主機名要與persistent對應(yīng)，要先設(shè)置個IP別名。Syndey#ifconfig enO aliasPerth#ifconfig enO alias3. 3 配置/etc/c I uster/rhosts 文件這個文件是的新增（必須）O每臺服務(wù)器都要修改修改完/etc/cIuster/rhosts 后用，refresh -s clcomd 命令更新配置3. 4準(zhǔn)備repository disk和共享存儲通過下面命令設(shè)競repository disk的PVID。集群主機都要設(shè)置。chdev - I hd）sk - a pv=yes通過下面方式設(shè)置共享存儲

10、（可以略過，集群配置好后通過C-SPOC設(shè)置）A主機：創(chuàng)迂卷如B主機：導(dǎo)入卷如4創(chuàng)建群集4.1初始化群集#smitty sysmi rror or smitty hacmpCIuster Nodes and Networks Initial Cluster Setup (Typical)Setup a Cluster, Nodes and NetworksSetup Cluster, Nodes and Networks (Typical) Type or select values in entry fields.Entry FieldsaustraliaperthSydneyPress E

11、nter AFTER making all desired changes.* Cluster NameNew Nodes (via selected cocmiunication paths) Currently Configured Node(s)4. 2 定義 Repository Disk如果上一步?jīng)]有自動創(chuàng)建，要手動創(chuàng)建。#smitty sysmi rror ?Cluster Nodes and Networks ?Initial Cluster Setup (Typical)Define Repository and Cluster IP AddressDefine Reposit

12、ory and Cluster IP AddressType or select values in entry fields.Press Enter AFTER making all desired changes.Entry Fields Cluster Name Repository DiskCluster IP Addressaustraliadisk2 «通過下拉選陰選擇磁盤+4. 3 配置 pers i stent i p#smit sysmi rrorCIuster Nodes and Networks ?Manage Nodes ?Conf i gure Pers i

13、 stent Node IP Labe I/Addresses?Add a Persistent Node IP Label/Address ?SeIect a Node net ether 01Entry FieldsSyndey選擇節(jié)點Syndey配競 Network Namenet_ether_01+ Node IP Label/AddressSyndeyNetmask(1Pv4)/Pref ix Length (IPv6)選擇節(jié)點Syndey配置Entry Fields Node NamePerth* Network Namenet_ether_Ol+ Node IP Label/Ad

14、dressPerth+Netmask (1 Pv4) /Pref i x Length (IPv6) Node Name4.4同步#smit sysmi rrorCluster Nodes and NetworksVer i fy and Synchronize CIuster ConfigurationVerify and Synchronize Cluster ConfigurationMove cursor to desired item and press Enter, use arrow keys to scroll.Any time you change the cluster c

15、onfiguration, those changes must be propagated to al 1 nodes in the cluster. PowerHA SystemMirror w perform a set of verification checks to make sure the configuration is viable.Depending on the nature of the changes, this operation can disrupt the applications currently under PowerHA SystemMirror c

16、ontrol.Press enter when you are ready to begin.Fl=HelpF2-RefreshF3=CancelF8=ImageF10=ExitEnter-Do/=Findn:Find NextARE YOU SURE?Continuing may delete Information you may want to keep. This is your last chance to stop before continuing.Press Enter to continue.Press Cancel to return to the application.

17、fBhg I pF2=Ref reshF 3=CancelF8=ImageF10=ExitEnter=Do4.5查看配置信息#PowerHA cluster topoIogy/usr/es/sb i n/cIuster/ut iIities/cltopinfo4.6配置資源和應(yīng)用程序4. 6.1 添加 App I i cat ions 腳本SyndeyPerth創(chuàng)建腳本目錄mkdi r /HA71/mkdir /HA71/創(chuàng)建腳本touch /HA71 /touch /HA71/touch /HA71/touch /HA71/修改腳本權(quán)限chmod +x /HA71 /chmod +x /HA

18、71 /chmod +x /HA71/chtnod +x /HA71/#smitty sysmi rrorCluster Applicat ions and ResourcesResourcesConfigure User Applicat ions (Scr i pts and Monitors)Applicat ion Control I er Scr i ptsAdd Application Controller ScriptsType or select values in entry fields.Press Enter AFTER making all desired change

19、s.Entry Fields Application Controller Namedbac Start Script/HA71/ Stop ScriptApplication Monitor Name(s)/HA71/+4. 6.2配置服務(wù)IP#smitty sysmi rrorCluster Applicat ions and ResourcesResourcesConf i gure Service IP Labe Is/AddressesAdd a Service IP Label/AddressNetwork NameII Bove cursor to desired item an

20、d press Enter.1II11I net_ether_01 (/24 /24)1下一步：選擇服務(wù)IPAdd a Service IP Labe I/AddressType or select values in entry fields.Press Enter AFTER making all desired changes.Entry Fields IP Label/AddressSyndey_svc +Netmask(1Pv4)/Pref ix Length (IPv6) Network Name口etherOI4. 6.3創(chuàng)建資源組#smitty sysmi rrorCluste

21、r Applicat ions and ResourcesResource Groups.Add a Resource GroupAdd a Resource GroupType or select values in entry fields.Press Enter AFTER making all desired changes.Entry Fields Resource Group Namedbrg Participating Nodes (Default Node Priority)Startup Pol icysydney perth +Online On Home Node 0&g

22、t; +Fai lover PolicyFai lover To Next Pr io> +FaiIback Pol icyNever Fai Iback > + <<主備推薦 nerver fa back4. 6.4修改資源組屬性#smitty sysmi rrorCluster Applicat ions and ResourcesResource Groups.Change/Show Resources and Attr ibutes for a Resource GroupChange/Show Al I Resources and Attributes for

23、 a Resource Group Type or select values in entry fields.Press Enter AFTER making all desired changes. TOPEntry FieldsResource Group NamedbrgParticipating Nodes (Default Node Priority)Startup Pol icySydney perthOnline On Home Node 0>Fai lover Pol icyFai lover To Next Pr io>FaiIback Pol icyFallb

24、ack To Higher Pr>Fai Iback Timer Pol icy (enpty is iirmediate) +Service IP Labels/AddressesSyndey-svc +Application Controllersdbac +Volume Groups+Use forced varyon of volume groups, if necessaryfalse+4. 6.5添加卷組#smitty sysmi rrorSystem Management (C-SPOC)StorageVo Iume GroupsCreate a Volume Group

25、«選擇節(jié)點名Node NamesMove cursor to desired item and press Esc+7.ONE OR MORE items can be selected.Press Enter AFTER making al I selections.> perth> Sydney下一步:Move cursor to desired item and press Esc+7.ONE OR MORE items can be selected.Press Enter AFTER making al I selections.00c1f170674f3d6b

26、 ( ndisk3 on all selected nodes ) <<選壞共享磁盤下一步:Create a Volume GroupType or select values in entry fields.Press Enter AFTER making all desired changes.TOPEntry FieldsNode NamesPerth, SydneyResource Group Namedbrg+PVID00c1f170674f3d6bVOLUME GROUP namedbvgPhysical partition SIZE in megabytes512+V

27、olume group MAJOR NUMBER37#Enable Cross-Site LVM Mirroring Verificationfalse+Enable Fast Disk Takeover or Concurrent AccessFast Disk Takeover+Volume Group TypeOriginalCRIT1CAL volume groupno+4. 6.6再次同步5 HACMP狀態(tài)管理5.1 啟停 HACMP啟動 smi t c I startStart Cluster ServicesType or select values in entry field

28、s.Press Enter AFTER making al 1 desired changes.Entry Fields Start now. on system restart or bothnow+Start Cluster Services on these nodesSyndey+ Manage Resource GroupsAutomatically+BROADCAST message at startupfalse+Startup Cluster Information Daemontrue+Ignore verification errorsfalse+Automatically

29、 correct errors found duringInteractively+cluster start停止 smit c 1 stopStop Cluster ServicesType or select values in entry fields.Press Enter AFTER making all desired changes.Entry Fields Stop now, on system restart or bothnow+Stop Cluster Services on these nodesSyndeyBROADCAST cluster shutdowntrue+

30、 Select an Action on Resource GroupsBring Resource Groups> +5.2 查看HACMP最終狀態(tài)Issrc -Is clstrmgrES|egrep "state|vers ion|vrmf|fi x*/usr/sb i n/1 sc Iuster -c #CAA cluster summary configuration/usr/sbin/lscluster -m #CAA cluster node status/usr/sbin/1 sc Iuster -i#CAA cluster interface status/us

31、r/es/sbin/cluster/utiIities/cIRGinfo -m tfCluster resource group status /usr/es/sb i n/c I uster/c I stat # 更看狀態(tài).三示不正,Q ':在次5.3 修復(fù)查看HACMP狀態(tài)不正常問題：執(zhí)行/usr/es/sbin/cluster/cIstat,顯示不正常rootpbfes2app01:/usr/es/sbin/cluster#./clstat Failed retrieving cluster information.There are a number of possible c

32、auses: cli nfoes or snmpd subsystems are not active, snmp i s unresponsive.snmp i5 not configured correctly.Cl lister servi ces are not active on any nodes.Refer to the hacmp Administration Guide for more Information.修復(fù)方法：在Syndey上執(zhí)行root&Syndey:/#stopsrc -s snmpdroot?Syndey: / #/u$r/sb in/sntrpv3 ssw -1root®