1、2018/4/12,目标,1.如何安装aix2.简单的aix系统概念文件系统逻辑卷管理设备管理3.备份，数据收集4.如何寻求IBM的帮助,2018/4/12,What is AIX,基于UNIX V操作系统改进的操作系统最容易上手的操作系统Linux亲和度,2018/4/12,P570前视图,2018/4/12,P570后视图,2018/4/12,P570机柜连接前视图,2018/4/12,P570机柜连接后视图,2018/4/12,系统开机,一.将外接电源与机柜电源接通；二.将机柜电源与主机电源接通，注意顺序为先外设，后主机，先扩展柜，后主机柜；三.上电完毕后按控制面板上的开关启动系统（本步
2、也可在系统中执行）；四.启动分区；,2018/4/12,系统关机,2018/4/12,AIX 系统安装, 磁带 CD-ROM 网络安装（NIM）,2018/4/12,系统安装步骤1,2018/4/12,系统安装步骤2,2018/4/12,系统安装步骤3,2018/4/12,系统安装步骤4,2018/4/12,系统安装步骤5,2018/4/12,系统安装步骤6,2018/4/12,系统安装步骤7,2018/4/12,系统安装步骤8,2018/4/12,系统安装步骤9,2018/4/12,系统安装步骤10,2018/4/12,系统安装步骤11,2018/4/12,系统安装步骤12,2018/4/1
3、2,系统安装步骤13,2018/4/12,系统安装步骤14,2018/4/12,系统安装步骤15,2018/4/12,系统安装步骤16,2018/4/12,系统安装步骤17,2018/4/12,系统安装步骤,2018/4/12,如何安装补丁,smitty installp 装机完要打的一些剩余的补丁包smitty update_all 5.2.07-5.2.08,2018/4/12,系统管理工具（SMIT）,2018/4/12,设备的管理和使用,物理设备物理设备是指以某种方式与计算机系统相连的实际硬件 设备驱动逻辑设备逻辑设备是用户或应用程序访问物理设备的界面，用户或应用程序通过逻辑设备来访问
4、物理设备 （在很多情况下会有一个物理设备对应多个逻辑设备）/dev 目录,2018/4/12,设备的管理和使用,块设备 是一种随机访问设备，用户可以随时访问设备中的任意一块 如：光驱、软驱、磁盘等字符设备 是一种原始的、面向字符流的顺序存取设备，每次按顺序读写一个字符 如：终端、网卡等,2018/4/12,AIX设备的层次,在AIX系统中，按类、子类、型号三个层次对设备进行分类 ：类(class)是按设备的功能划分的，例如打印机类、硬盘类、适配卡类等。子类(subclass)是按设备的某种相似性划分的，例如适配卡类可分为PCI子类和MCA子类。型号(type)是按设备的机型或生产厂家划分的，例
5、如磁带机可分为4mm12gb、 8mm20gb等型号。,2018/4/12,AIX设备状态,在AIX系统中，设备可以处在下列三种状态之一 ：Undefined未定义状态如果设备处于此状态，则系统没有该设备的任何信息。Defined已定义状态处于此状态，说明系统已经存有该设备的全部信息，并已经给设备分配好逻辑设备名和端口，但当前设备还不能被使用。Available可用状态处于此状态的设备能被正常使用。 #cfgmgr v识别设备,AIX 存储管理介绍(LVM),文件 目录 文件系统 逻辑卷 卷组 物理卷（物理存储）,2018/4/12,AIX 存储管理介绍(LVM),2018/4/12,IBM/
6、AIX Disk Terminology,The AIX disk terminology includesLVM (Logical Volume Manager)PV 物理卷，通常=硬盘VG 卷组PP 物理分区LV 逻辑分区JFS 日志文件系统,2018/4/12,Logical Volume Manager = LVMPart of AIXno feenot optionalactually very goodManages disks and their attributesby commandsthrough database of configurationthrough devic
7、e driversfeatures like mirroring, RAID, allocation,LVM,2018/4/12,PV,Physical Volume = PVLogical device, map to a physical disk OR a RAID OR a Shark . Named by AIX as hdisk0, hdisk1, hdisk2, .Regardless of the technologySCSISSARAID5 subsystemFibre channel/SAN,hdisk0 hdisk1 hdisk2,Section 6.3,2018/4/1
8、2,VG,Volume Group = VGA group of disksTo use a disk it must be added to a VG first.Group operations:Disk space always allocated within single VGexportvg and importvgvaryonvg and varyoffvg - allows High Availability HACMPFirst VG called rootvg,volume group A,volume group B,Section 6.4,2018/4/12,rootv
9、g,Root Volume Group - rootvgCreated automatically while installing AIXPlaced within this VGAIX files Initial paging and other LVsUsually only the first disk Or two to allow mirroring of rootvgOften internal disksRecommendKeep to a small number of disks (see mksysb later),volume group rootvg,volume g
10、roup B,2018/4/12,Other VG,Other Volume GroupsOther VG created by the System Admin.Name typically related to their useOptionally ending in vgVG has one or more disksCan later add a disk to a VGCan remove disk from VG - if empty,volume group rootvg,volume group B,2018/4/12,VG,Volume GroupsDisks within
11、 VG cut into chunks calledPhysical Partitions (PP)Minimum allocation unitTypically, 8, 16 or 32 MB (older 2GB disks = 4 MB)Each VG fixed PP size.Volume group has total, free and used PPs (lsvg)Disks do not need to be the same size oreven the same technology,volume group rootvg,volume group B,2018/4/
12、12,LV,Logical Volume - 1It is a large piece of diskCan be used for lots of tasks likeraw device or file system or paging spaceCreated within single VGLV has many attributes like:can be one disk or more disks or all disks of VGmirrored or stripedparticular part of disk (physical location)size etc.,vo
13、lume group B,Section 6.5,2018/4/12,LV,Logical Volume - 2Sizes is in multiple Physical PartitionsPP = 16 MB and LV size of 64 PPs = 1 GBPPs allocated by policiesmin or max spread across disksmin=fill disk then move on, max=evenly spreaddisk placement: center, middle, edge, etc.mirrors via extra copie
14、s, up to 3stripes via stripe widthMWC and others,volume group B,2018/4/12,LV,Logical Volume - 3Can be grown in sizeCan be removedCan be moved to different disks within VGCan later have a filesystem created in itAttributes can be changed like:mirror added or removed or policy changed,volume group B,2
15、018/4/12,LP vs PP,LV is constructed by LPsLPs is map to PPs (1:1, 1:2 OR 1:3)The size: LP=PPThe maped PPs can be in different disks.,volume group B,2018/4/12,LVM Command,How can we viewlsvglspvlslv,Volume Groups,Physical Volume,Logical Volume,Section 6.8,2018/4/12,LVM lsvg Command,lsvgLists volume g
16、roup names: lsvgList VG details: lsvg List LVs within VG: lsvg -l List PV within VG: lsvg -p ,Volume Groups,Physical Volume,Logical Volume,2018/4/12,LVM lspv Command,lspvLists physical group names: lspvList PV details: lspv hdiskNList LVs within PV: lspv -l hdiskNList PP within PV: lspv -p hdiskN,Vo
17、lume Groups,Physical Volume,Logical Volume,2018/4/12,LVM lslv Command,lslvLists LV names: lsvg -l List LV details: lslv List LVs within PV:lslv -l List PP within PV: lslv -p Show mirrored LV: lslv -m ,Volume Groups,Physical Volume,Logical Volume,2018/4/12,文件和目录,文件是UNIX系统的基本单位，共有三种不同类型的文件普通文件 普通文件就是通
18、常所指的程序文件和文本文件，用来存储数据。它们既可是文本的，也可是二进制的，有的还可以执行。目录文件 目录文件包括了用来存取其它文件的信息，简称为目录。特殊文件 一个特殊文件定义了一个先进先出(FIFO)的管道或者一个物理设备，通常指磁带、磁盘等外部设备。,2018/4/12,什么是文件系统,2018/4/12,AIX标准文件系统,2018/4/12,系统自动创建的LV,Paging Space,需要Paging Space的原因,Paging Space的计算,If real memory256M:paging space=512M+(real mem-256)X1.25Paging spa
19、ce should be no more than 20% of total disk space,2018/4/12,备份与恢复,remember what smitty mksysb will do: a. Backup operating system configuration b. Backup only data on rootvg c. Backup only mounted filesystemsDifferent machine backup and restoreAIX OS cdrom maybe needed.,系统备份,2018/4/12,备份与恢复,2018/4/1
20、2,备份与恢复,2018/4/12,备份与恢复,2018/4/12,系统错误日志,# errpt -dH 仅列出硬件错误日志# errpt -dS仅列出软件错误日志# errpt -a 列出所有错误日志地详细信息# errpt -aj errorID 列出error ID对应地详细信息,作为系统管理员，至少2到3天应该检查一次系统地错误报告：在主机上使用errpt指令来检查：,系统错误日志,如果错误日志中已经没有有效地信息，可以将其进行清除： # errclear 0 删除所有日志# errclear 1删除所有1天前的日志如果日志中依然有一些有用的信息，那么在删除日志前，可以将日志保存到文件
21、中：# errpt /tmp/log20021031.list保留所有日志# errpt -a /tmp/log20021031.out保留所有日志的详细信息 注意同时还要检查系统发给ROOT的MAIL# mail,Write down all code on the LED,4 or 8 digits can be seen during system boot8xxx, 9xxx, Axxx, Bxxx are service processor checkpoint codes; Exxx are firmware checkpoint codes usually, nothing is
22、 on the screen after successful system boot Record each code you see when the system is abnormalfor example, you will see: 0c0, which means system had experienced a dump,How to keep system dump,some time the system will experienced dump or just hang(no any response from display, keyboard, mouse and
23、network), it is important to keep system dump for IBM to analyze the reason. Failed to get the dump may means that IBM cant find out the reason. Use smitty dump to set related preference: 1. Primary dump device 2. Dump copy directory 3. Estimate dump size 4. Always allow dump let us see more detail
24、information in AIX4.3 trouble shooting guide,Send useful messages to IBM,collect as more information as you can 1. 4 or 8 digit error code on LED 2. What you are doing as that time 3. What had been changed? 4. Use snap -gc is a very simple way to collect system HW information and error logs, when co
25、mmand finished /tmp/ibmsupt/snap.pax.Z 5. If there is a system dump, you should usesnap -gfkDc6. Then, send the file to IBM by email or our engineer will get it on site,Snap gc （收集硬件相关）snap ac（收集软件相关）Snap ec (收集hacmp相关）,What else?.（收集数据）,2018/4/12,What else?.,别忘记了IBM的客户服务热线主机硬件故障报修电话：（7*24小时*365天）800-810-6677 系统支持及咨询热线：（5*8小时*工作日）800-810-1818转5200,2018/4/12,Q & A,
