Solaris
Upcoming SlideShare
Loading in...5
×
 

Solaris

on

  • 844 views

Tips For Solaris

Tips For Solaris

Statistics

Views

Total Views
844
Views on SlideShare
844
Embed Views
0

Actions

Likes
0
Downloads
71
Comments
0

0 Embeds 0

No embeds

Accessibility

Categories

Upload Details

Uploaded via as Microsoft Word

Usage Rights

© All Rights Reserved

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Processing…
Post Comment
Edit your comment

Solaris Solaris Document Transcript

  • Akadia Information TechnologySUN Free SoftwarePCNFS installierenInstallation Solaris mit Openwindows (Grafikkarte)How to Backup a SystemPackages (Software die installiert wurde)Monitor Mode (OK Prompt)Kernel AnalyseDefaults einstellenWichtige KonfigurationsfilesAdmin KommandosLAN konfigurierenIP-Routing konfigurierenDNS konfigurierenAnonymous FTP aufsetzenNFS-Client KonfigurationNFS-Server KonfigurationAutomounterModem konfigurierenSCSI-Harddisk an SUN Hardware anschliessenList Solaris Hardware ConfigurationShow Swap Space currently installedShow Operating System Patch LevelHow to install a Sun Solaris Jumbo Patch ?Tracing System CallsTroubleshooting Solaris Device FilesShort Tips to maintain Sun SolarisIP-Aliasing for SUN SolarisSolaris automounter installs filesystems by default in /netSolaris keyboard utilityMonitoring PerformanceEnable file system journaling on Solaris 7 and 8Solaris Syslog Daemon DebuggingDoes each Oracle Process use more than 100M memory ?Sizing up Solaris Memory with the RMCmem PackageUsing Sun Solaris Manuals directly from CD-ROMWhy is the Sun Solaris System Corefile helpful ?DLT-TAPE UNIT INSTALLATION on Solaris 7/8/9Reconfigure Devices on SolarisOpenBoot DiagnosticsWhy doesnt my .forward file workSimple Shell Script to backup your Files
  • SUN Free SoftwareUnter http://www.sunfreeware.com findet man "ready to use" Software für SUN Solaris, wie beispielsweise TOP, AGCC, GDB etc. Download via FTP von: «ftp://nce.sun.ch/pub/freeware/sparc/7» PCNFS installierenCD-ROM Solaris Intranet Extension (siehe auch Solaris Server Intranet Extension Installation)$ su$ cd /cdrom/cdrom0/nfsc/sparc$ pkgadd -d `pwd` Installation Solaris mit Openwindows (Grafikkarte) Hostname$ uname -u Network Interface$ ifconfig -a /etc/hosts, /etc/netmasks definieren Static IP-routes definieren/etc/rc2.d/S79staticroutes CD-ROM rausnehmen$ eject cdrom Disklayout kontrollieren$ prtvtoc /dev/rdsk/.... Automounter konfigurieren/etc/auto_master, /etc/auto_home Device File für DAT/dev/rmt/0l (tar cvf /dev/rmt/0l)
  • /etc/system definieren konfigurieren (Prestoserve, Oracle, Transtec) Logfile der Installation: /var/sadm/system/logs/install_log Installation über serielles Terminal an Nullmodem KabelOK boot cdrom - w (Terminal an ttya) How to Backup a System$ init 0OK boot -s$ fsck -m /dev/dsk/c0t0d0s0 (und übrige Filesysteme)$ tar cvf /dev/rmt/0l Packages (Software die installiert wurde)Anzeige der installierten Packages$ pkginfoCheck ob Package SUNWpcnfd correct installiert ist$ pkgchk -v SUNWpcnfdPackage installieren (Path ist meistens /cdrom/cdrom0/....)$ pkgadd -d <Path to CDROM> SUNWpcnfdDas Package SUNWpcnfd entfernen$ pkgrm SUNWpcnfd Monitor Mode (OK Prompt)In single user mode bootenOK boot -sKernel zwingen /devices neu aufzubauen nach dem Anschluss von neuer HardwareOK boot -rDetaillierter Bootvorgang
  • OK boot -vVom CD-ROM aus booten: Notboot !OK boot cdromAngeschlossene SCSI-Geräte testenOK probe-scsiList System Devices, e.g. SUNW,hme = Sun Fast Ethernet PCI AdapterOK show-devsList Network DevicesOK show-netsMonitoring Network ActivityOK apply watch-net <full path name of the hme interface, see show-devs>Monitor Variablen ändern, anzeigenOK eepromOK eeprom ttya-mode=38400,8,n,1,h Kernel AnalyseWelche Kernel-Module sind geladen ?$ modinfoKernel Konfiguration/etc/systemLogfile von syslog/var/adm/messagesKonfiguration des syslog Daemon/etc/syslog.conf Defaults einstellen
  • Directory mit Default files/etc/defaultRemote root logins erlauben/etc/default/loginTimezone setzen/etc/default/init Wichtige KonfigurationsfilesMaster-File beim Booten/etc/inittabRun-Level Start/Stop Files/etc/rc?.dScripts für Run-Levels/etc/init.d Admin Kommandos$ shutdown -g0 -i0$ reboot (entspricht init 6) LAN konfigurierenKonfiguration der LAN-Interfaces$ ifconfig -a Netmask setzen: siehe /etc/netmasks Jedes LAN-Interface hat /etc/hostname.le0 mit Hostnamen LAN-Setup: /etc/rcS.d/S30rootusr.sh (Interfaces konfigurieren) /etc/rc2.d/S72inetsvc (LAN konfigurieren)Phys Addressen nachschauen$ arp -aNet to Media Table
  • Device IP Address Mask Flags Phys Addr------ -------------------- --------------- ----- ---------------le0 rabbit 255.255.255.255 00:60:08:57:17:86le0 quorum 255.255.255.255 SP 08:00:20:89:27:03le0 arkum 255.255.255.255 00:a0:24:4b:60:1c IP-Routing konfigurierenAlle hosts im Netz 193.72.239.0 werden über den Router 193.72.194.201 erreicht.$ route add net 193.72.239.0 193.72.194.201 1Der host 146.228.14.10 wird über den Router 193.72.194.100 erreicht. Siehe File /etc/rc2.d/S79staticroutes.$ route add host 146.228.14.10 193.72.194.100 1Routing Tabelle kontrollieren$ netstat -nr DNS konfigurierenAngabe des DNS Nameservers/etc/resolv.confReihenfolge definieren/etc/nsswitch.conf Anonymous FTP aufsetzenSiehe Solris2 Administration Seite 103 und ff NFS-Client Konfiguration/etc/vfstab (Soll) --> /etc/mnttab (Ist)mount -F nfs -o bg,ro,soft gondwana:/usr/software /softwareNFS-Server wird in /etc/init.d/nfs.client start gestartet.Anzeige welche Directories gondwana zum mounten freigegeben hat
  • dfshares gondwanaRESOURCE SERVER ACCESS TRANSPORTgondwana:/export/home/zahn gondwana - -gondwana:/export/home/steiner gondwana - - NFS-Server Konfiguration/etc/dfs/dfstab (Soll) --> /etc/dfs/sharetabDirectory read-only freigeben$ share -o ro /usr/softwareAlle Directories in /etc/dfs/dfstab freigeben$ shareallAlle Directories in /etc/dfs/dfstab zurücknehmen$ unshareallNFS-Server wird gestartet in/etc/init.d/nfs.serverAnzeige der freigegbenen lokalen Direcories$ shareAnzeige welche Clients nutzen welche Directories eines NFS-Servers$ dfmounts -F nfs gondwanaRESOURCE SERVER PATHNAME CLIENTSgondwana /export/home/zahn paragon.glue.ch,rabbit.glue.ch Automounter- /etc/auto_master (Master Map konfigurieren)- /etc/auto_home (Home Direcories verwalten)- autofs ist ein spezielles Filesystem- automount -v (Nach einer Aenderung an einer Map ausführen) Modem konfigurieren
  • Siehe spezielles Dokument SCSI-Harddisk an SUN Hardware anschliessenBeispiel: SCSI-Disk Seagate ST150176L, 50MB an SUN Ultra Enterprise 1 Eintrag in /etc/format.dat vornehmen (Angaben von Lieferanten)disk_type = "Seagate ST150176L" : ctlr = "SCSI" : ncyl = 12022 : acyl = 2 : pcyl = 12024 : nhead = 22 : nsect = 369 : rpm = 7200 : bpt = 188928 Eintrag /etc/system für Solaris-2 Kernel, System booten** SCSI-Disc Konfiguration*set scsi_options=0x20 Disk anschliessen, SCSI-Adresse kontrollieren, Terminierung Unbedingt kontrolieren, dass eine SCSI-Adresse nicht mehrfach belegt ist. Dazu kann meistens hinte Tippschalter eingestellt werden. Man beacht, dass in der Regel das letzte Gerät terminiert werden m Disk formatieren (nur wenn notwendig !) In der Regel muss eine Disk nicht neu formatiert werden, ist dies jedoch notwendig so steht unter So format zur Verfügung.formatAVAILABLE DISK SELECTIONS:0. c0t0d0 <SUN2.1G cyl 2733 alt 2 hd 19 sec 80> /sbus@1f,0/espdma@e,8400000/esp@e,8800000/sd@0,01. c0t1d0 <SUN2.1G cyl 2733 alt 2 hd 19 sec 80> /sbus@1f,0/espdma@e,8400000/esp@e,8800000/sd@1,02. c0t2d0 <IBM-DDRS-39130-S71D cyl 8186 alt 2 hd 10 sec 218> /sbus@1f,0/espdma@e,8400000/esp@e,8800000/sd@2,03. c0t4d0 <SEAGATE-ST118273N-5764 cyl 7499 alt 2 hd 20 sec 237> /sbus@1f,0/espdma@e,8400000/esp@e,8800000/sd@4,04. c0t5d0 <SEAGATE-ST150176LW-0002 cyl 12022 alt 2 hd 22 sec 369> /sbus@1f,0/espdma@e,8400000/esp@e,8800000/sd@5,0Specify disk (enter its number): 4format> typeAVAILABLE DRIVE TYPES:
  • 0. Auto configure1. DDRS-391302. Seagate ST1182733. Seagate ST150176L4. Quantum ProDrive 80S5. Quantum ProDrive 105S6. CDC Wren IV 94171-3447. SUN01048. SUN02079. SUN032710. SUN034011. SUN042412. SUN053513. SUN066914. SUN1.0G15. SUN1.0516. SUN1.3G17. SUN2.1G18. SUN2.9G19. IBM-DDRS-39130-S71D20. SEAGATE-ST118273N-576421. SEAGATE-ST150176LW-000222. otherSpecify disk type (enter its number)[21]: 21format> format (confirm with "yes") Disk partitionierenDadurch wird die Disk in logische Teile unterteilt. Jeder teil enthält ein eigenes Filesystem.format> partNun die Partitionierungsdaten eingeben, zBpartition> printCurrent partition table (original):Total disk cylinders available: 2733 + 2 (reserved cylinders)Part Tag Flag Cylinders Size Blocks0 root wm 0 - 204 152.15MB (205/0/0) 3116001 swap wu 205 - 377 128.40MB (173/0/0) 2629602 backup wm 0 - 2732 1.98GB (2733/0/0) 41541603 home wm 378 - 1017 475.00MB (640/0/0) 9728004 unassigned wm 0 0 (0/0/0) 05 unassigned wm 1018 - 1928 676.13MB (911/0/0) 13847206 usr wm 1929 - 2732 596.72MB (804/0/0) 12220807 unassigned wm 0 0 (0/0/0) 0
  • Label erzeugen (aktuelle Partitionierung speichern)partition> labelpartition> ypartition> quitformat> quit Filesystem erstellennewfs -v -m 0 /dev/rdsk/c0t5d0s0Damit wird ein Filesystem mit 0 % Min-Free auf der Partition 0 der Disk an der SCSI-Adresse 5 erstellt. Filesystem mountenDazu den folgenden Eintrag in /etc/vfstab vornehmen#device device mount FS fsck mount mount#to mount to fsck point type pass at boot options/dev/dsk/c0t5d0s0 /dev/rdsk/c0t5d0s0 /u02 ufs 6 yes - List Solaris Hardware Configuration$ /usr/sbin/prtconf Show Swap Space currently installedMultiply the Blocks column by 512$ swap -lswapfile dev swaplo blocks free/dev/dsk/c0t0d0s1 32,1 16 262944 262944262944 * 512 = 134 MB Show Operating System Patch Level$ showrev -pPatch: 105181-16Note, that Patchlevel 105181-15 is minimal needed for Oracle 8.1.6
  • How to install a Sun Solaris Jumbo Patch ?- Download the Patch from: http://sunsolve.sun.com- Read the README File included in the Patch- Usually the only thing you have to do is:$ cd <patch cluster directory>$ ./install_custer$ cat /var/sadm/install_data/<luster name>_log$ showrev -pReboot the system Tracing System CallsYou can trace system calls with truss on Solaris an strace on Linux$ truss svrmgrl Troubleshooting Solaris Device FilesIf you suspect troubles with your Solaris device files, e.g. system doesnt boot after a filesystem check, you may repsystem using the following commands.Halt the system immediately with the keys STOP-A, you will now see the boot prompt: OKSTOP-AReset the machine withOK resetBoot the machine withOK boot -rThe command boot -r will rebuild all devices files according to your attached hardware. If you cannot boot the mactry the following commands: drvconfig, disks, tapesdrvconfig - configure the /devices directoryThe default operation of drvconfig is to create the /devices directory tree that describes, in the filesystem namespaclayout of a particular machine. Hardware devices present on the machine and powered on as well as pseudo-driversunder /devices. Normally this command is run automatically after a new driver has been installed (with add_drv(1Msystem has been rebooted.
  • disks - creates /dev entries for hard disks attached to the systemDisks creates symbolic links in the /dev/dsk and /dev/rdsk directories pointing to the actual disk device special files/devices directory tree.tapes - creates /dev entries for tape drives attached to the systemTapes creates symbolic links in the /dev/rmt directory to the actual tape device special files under the /devices direcsearches the kernel device tree to see what tape devices are attached to the system. Short Tips to maintain Sun SolarisHere are some short tips for common tasks on SUN Solaris 2.6, 7 and 8Important SUN Solaris Commands$ who -r # Show Run Level$ /usr/sbin/prtconf # Print the complete system configuration$ /sbin/mountall -l # Mount all local filesystems.$ /sbin/init S # Changing to single user modeShow currently mounted filesystems# /etc/mnttab: Contains information about devices that# are currently mounted. If there are mounted filesystems# with quotas enabled, display themif /usr/bin/cut -f 4 /etc/mnttab | /usr/bin/egrep ^quota|,quota >/dev/null 2>&1; then echo There are mounted filesystems with quotas enabledfiHow to enable system activity data gathering# You will also need to uncomment the sa entries in# the system crontab /var/spool/cron/crontabs/sys.# Refer to the sar(1) and sadc(1m) man pages# for more information.$ /usr/bin/su sys -c "/usr/lib/sa/sadc /var/adm/sa/sa`date +%d`"How a new, unused Solaris system is setup ?# sysidtool is a suite of five programs that configure a new# system, or one that has been unconfigured with sys-# unconfig(1M). The sysidtool programs run automatically at# system installation, or during the first boot after a# machine has been successfully unconfigured.
  • ## These programs have no effect except at such times, and# should never be run manually.# System Files arecat /etc/nodenamecat /etc/hostname.*cat /etc/default/initcat /etc/defaultdomaincat /etc/inet/hostscat /etc/inet/netmasksHow to configure Asynchronous PPP ?Configure /etc/asppp.cf for the aspppd daemon$ /usr/sbin/aspppd -d 1How to get and set TCP/IP driver configuration parameters ?# Getting Parameters Supported By The TCP Driver# To see which parameters are supported by the TCP driver,# use the following command:$ ndd /dev/tcp ?# The following command sets the value of the parameter# ip_forwarding in the IP driver to zero. This disables IP# packet forwarding.Disable IP Forwarding$ /usr/sbin/ndd -set /dev/ip ip_forwarding 0Enable IP Forwarding (Machine acting as a Router)$ /usr/sbin/ndd -set /dev/ip ip_forwarding 1How to set Default Route on Solaris ?# Configure default routers using the local "/etc/defaultrouter"# configuration file. The file can contain the hostnames or IP# addresses of one or more default routers.## The default routes listed in the "/etc/defaultrouter" file will# replace those added by the kernel during diskless booting. An# empty "/etc/defaultrouter" file will cause the default route# added by the kernel to be deleted.
  • ## Note that the default router file is ignored if we received routes# from a DHCP server. Our policy is to always trust DHCP over local# administration.# Set Default Route$ route -n add default <Default Route from /etc/defaultrouter># Show Default Route$ /usr/sbin/route -fndefault 128.128.128.11 doneHow to set NIS domainname if locally configured ?if [ -f /etc/defaultdomain ]; then /usr/bin/domainname `cat /etc/defaultdomain` echo "NIS domainname is `/usr/bin/domainname`"fiRPC (Remote Procedure Call) Configuration# rpcbind - universal addresses to RPC program number mapper# rpcinfo - report RPC informationSolaris Keyserv Daemon# keyserv is a daemon that is used for storing the private# encryption keys of each user logged into the system. These# encryption keys are used for accessing secure network ser-# vices such as secure NFS and NIS+.$ /usr/sbin/keyservHow to start the Solaris DNS server "in.named"# If this machine is configured to be an Internet# Domain Name System (DNS) server, run the name daemon.# Start named prior to: route add net host,# to avoid dns gethostbyname timout delay for# nameserver during boot.if [ -f /usr/sbin/in.named -a -f /etc/named.conf ]; then echo starting internet domain name server. /usr/sbin/in.named &fi
  • Where to find syslogd messages ?Configuration File: /etc/syslog.confMessage File: /var/adm/messages IP-Aliasing for SUN Solaris# How to setup IP-Alias on SUN Solaris1. Setup File /etc/hostname.hme0:1 for the second IP-Address cat /etc/hostname.hme0:1 ldap2. Insert IP-Address in /etc/hosts # # Internet host table # 128.128.128.11 ux-portal1 # IP-address on hme0:0 128.128.128.20 ldap # IP-alias on hme0:13. Start alias IP-Address on Interface in /etc/rc2.d S99ipalias -> ../init.d/ipalias #!/bin/sh # Akadia AG, Arvenweg 4, CH-3604 Thun # ---------------------------------------------------------------------- # File: ipalias # # Autor: Martin Zahn / 10.05.2000 # # Purpose: Setup second IP address on hme0:1 # ---------------------------------------------------------------------- if [ -f /etc/hostname.hme0:1 ] then case "$1" in start) # Start second IP address on hme0:1 echo "Start multi-homed server for UX-ALIAS1 on hme0:1" ifconfig hme0:1 128.128.128.20 up ;;
  • stop) # Stop second IP address on hme0:1 echo "Stop multi-homed server for UX-ALIAS1 on hme0:1" ifconfig hme0:1 128.128.128.20 down ;; esac fi4. Check IP-Address on second Interface ifconfig -a Solaris automounter installs filesystems by default inThe Solaris automount utility installs autofs mount points and associates an automount map with each mount poautofs file system monitors attempts to access directories within it and notifies the automountd daemon. The dthe map to locate a file system, which it then mounts at the point of reference within the autofs file system. Youmap to an autofs mount using an entry in the /etc/auto_master map or a direct map in /etc/auto_dirsystem is not accessed within an appropriate interval (five minutes by default), the automountd daemon unmounsystem.Default Mapping under /netThe mount point /net is by default the location, where automountd mounts NFS filesystems, which are exportemachines. Lets suppose, that you have the filesystem /home exported on the NFS server saphir, then the (Solariwith an active automounter will automatically mount this NFS filesystem under /net/saphir/.Mapping using /etc/auto_directYou probably doesnt want this default behavior. If you insert the following entry in /etc/auto_direct ..../opt/local -rw remote_machine:/local.... then, the directory /local on the remote machine "remote_machine" will be mounted on the local machine un/opt/local. Solaris keyboard utilityThe Solaris utility kbd manipulates the state of the keyboard or display the type of keyboard or change the defaultabort sequence effect. Suppose, that you do not want that everybody can halt the the system you must change the dWe also noticed, that the Solaris machines attached to a switch box, using a character terminal on a serial line, mayswitch from one machine to the other.SYNOPSISkbd [ -r ] [ -t ] [ -c on|off ] [ -a enable|disable ] [ -d keyboard device ]
  • kbd -i [ -d keyboard device ]DESCRIPTIONkbd manipulates the state of the keyboard, or displays the keyboard type or allows the default keyboard abort seqube changed. The default keyboard device being set is /dev/kbd.The -i option reads and processes default values for the keyclick and keyboard abort settings from the keyboard de/etc/default/kbd. Only keyboards that support a clicker respond to the -c option. If you want to turn clickingadd or change the current value of the KEYCLICK variable to the value on in the keyboard default file, /etc/defas shown here.KEYCLICK=onThen, run the command kbd -i to change the current setting. Valid settings for this variable are the values onvalues are ignored. If the variable is not specified in the default file, the setting is unchanged.The keyboard abort sequence (L1-A or STOP-A) on the keyboard and BREAK on the serial console input device osystems) effect may only be changed by the superuser, using the-a option. On most systems, the default effect of the keyboard abort sequence is to suspend the operating system andebugger or the monitor.If you want to permanently change the software default effect of the keyboard abort sequence, you can add or changvalue of the KEYBOARD_ABORT variable to thevalue disable in the keyboard default file, /etc/default/kbd, as shown here.KEYBOARD_ABORT=disableThen, run the command kbd -i to change the current setting. Valid settings for this value are the values enableOther values are ignored. If the variable is not specified in the default file, the setting is unchanged.OPTIONS -i Set keyboard defaults from the keyboard default file. This option is mutually exclusive with all other options except for the -d keyboard device option. This option instructs the keyboard command to read a process keyclick and keyboard abort default values from the /etc/default/kbd file. This option c only be used by the superuser. -r Reset the keyboard as if power-up. -t Return the type of the keyboard being used. -c On/Off state Turn the clicking of the keyboard on or off. -a Enable/Disable state; Enable or disable the keyboard abort sequence effect. Monitoring PerformanceThis chapter describes procedures for monitoring system performance by using the vmstat, iostat, df, andcommands. This is a list of the step-by-step instructions in this chapter.
  • How to Display Virtual Memory Statistics (vmstat)The following example shows the vmstat display of statistics gathered at five-second intervals.$ vmstat 5 procs memory page disk faults cpur b w swap free re mf pi po fr de sr f0 s3 -- -- in sy cs us sy id0 0 8 28312 668 0 9 2 0 1 0 0 0 1 0 0 10 61 82 1 2 970 0 3 31940 248 0 10 20 0 26 0 27 0 4 0 0 53 189 191 6 6 880 0 3 32080 288 3 19 49 6 26 0 15 0 9 0 0 75 415 277 6 15 790 0 3 32080 256 0 26 20 6 21 0 12 1 6 0 0 163 110 138 1 3 960 1 3 32060 256 3 45 52 28 61 0 27 5 12 0 0 195 191 223 7 11 820 0 3 32056 260 0 1 0 0 0 0 0 0 0 0 0 4 52 84 0 1 99Category Field Description Nameprocs Reports the following states: r The number of kernel threads in the dispatch queue b Blocked kernel threads waiting for resources w Swapped out LWPs waiting for processing resources to finishmemory Reports on usage of real and virtual memory: swap Available swap space free Size of the free listpage Reports on page faults and paging activity, in units per second: re Pages reclaimed mf Minor and major faults pi Kbytes paged in po Kbytes paged out fr Kbytes freed de Anticipated memory needed by recently swapped-in processes sr Pages scanned by page daemon (not currently in use). If sr does not equal zero, the page daemon has been running.disk Reports the number of disk operations per second, showing data on up to four disksfaults Reports the trap/interrupt rates (per second): in Interrupts per second sy System calls per second cs CPU context switch ratecpu Reports on the use of CPU time: us User time
  • sy System time id Idle timeHow to Display System Event InformationRun vmstat -s to show the total of various system events that have taken place since the system was last booted 0 swap ins 0 swap outs 0 pages swapped in 0 pages swapped out409376480 total address trans. faults taken 3075036 page ins 2601555 page outs 3812452 pages paged in 6525552 pages paged out 11007609 total reclaims 10927650 reclaims from free list 0 micro (hat) faults409376480 minor (as) faults 2957386 major faults102738273 copy-on-write faults 61711047 zero fill page faults1002562077 pages examined by the clock daemon 7881 revolutions of the clock hand 16716370 pages freed by the clock daemon 4999048 forks 1138206 vforks 5747009 execs741660225 cpu context switches736047593 device interrupts528054538 traps2496638575 system calls430283487 total name lookups (cache hits 95%) 81727 toolong 10484677 user cpu 9528364 system cpu443762786 idle cpu 16281790 wait cpuHow to Display Swapping StatisticsRun vmstat -S to show swapping statistics.procs memory page disk faults cpur b w swap free si so pi po fr de sr m1 m3 m4 m5 in sy cs us sy id0 0 0 8512 888 0 0 12 21 55 0 417 1 0 0 0 206 1040 308 2 2 96si = Average number of LWPs swapped in per second
  • so = Number of whole processes swapped outHow to Display Disk Utilization Information (iostat)You can display disk activity information by using the iostat command with a time interval. The following examstatistics gathered every five seconds.iostat 5 tty md1 md3 md4 md5 cputin tout kps tps serv kps tps serv kps tps serv kps tps serv us sy wt id 0 2 10 1 28 2 0 22 0 0 0 1 0 10 2 2 3 92 0 47 58 7 39 16 2 34 0 0 0 0 0 0 0 2 19 78 0 16 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 98 0 16 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 99 0 16 2 0 22 0 0 0 0 0 0 0 0 0 2 3 1 95 0 24 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 98For Each ... Field Name DescriptionTerminal tin Number of characters in the terminal input queue tout Number of characters in the terminal output queueDisk bps Blocks per second tps Transactions per second serv Average service time, in millisecondsCPU us In user mode sy In system mode wt Waiting for I/O id IdleHow to Display Extended Disk StatisticsRun iostat -xtc to get extended disk statistics. This command displays a line of output for each disk. extended device statistics tty cpudevice r/s w/s kr/s kw/s wait actv svc_t %w %b tin tout us sy wt idmd1 0.4 0.9 3.6 6.9 0.0 0.0 27.7 1 1 0 2 2 2 3 92md3 0.1 0.2 1.0 1.3 0.0 0.0 21.7 0 0md4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0 0md5 0.0 0.0 0.7 0.0 0.0 0.0 9.9 0 0md8 0.8 0.3 6.7 14.2 0.0 0.0 13.1 0 1md10 0.2 0.9 1.8 6.8 0.0 0.0 15.5 0 1
  • md11 0.2 0.9 1.8 6.8 0.0 0.0 14.8 0 1md30 0.0 0.2 0.5 1.3 0.0 0.0 11.4 0 0md31 0.0 0.2 0.5 1.3 0.0 0.0 10.2 0 0md40 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0 0md41 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0 0md50 0.0 0.0 0.4 0.0 0.0 0.0 9.4 0 0md51 0.0 0.0 0.4 0.0 0.0 0.0 7.3 0 0md80 0.4 0.3 3.3 14.2 0.0 0.0 10.3 0 0md81 0.4 0.3 3.3 14.2 0.0 0.0 11.7 0 1sd0 0.6 2.1 6.0 22.8 0.0 0.0 16.3 0 3sd1 0.6 2.1 6.0 22.8 0.0 0.0 15.2 0 2Field Name Descriptionr/s Reads per secondw/s Writes per secondKr/s Kbytes read per secondKw/s Kbytes written per secondwait Average number of transactions waiting for service (queue length)actv Average number of transactions actively being servicedsvc_t Average service time, in milliseconds%w Percentage of time the queue is not empty%b Percentage of time the disk is busyHow to Check CPU Utilization (sar)Display CPU utilization with the sar -u command. (The sar command without any options is equivalent to sar -u.) Amoment, the processor is either busy or idle. When busy, the processor is in either user or system mode. When idle,is either waiting for I/O completion or "sitting still" with no work to do.Measure CPU utilization during 5 secs one time.sar -u 5 1Measure CPU utilization during 60 secs 1440 times and write result in file sar.log.sar -u -o sar.log 60 1440To later review disk and tape activity from that period:sar -d -f sar.logField Name Description%sys Lists the percentage of time that the processor is in system mode%user Lists the percentage of time that the processor is in user mode
  • %wio Lists the percentage of time the processor is idle and waiting for I/O completion%idle Lists the percentage of time the processor is idle and is not waiting for I/OA high %wio generally means a disk slowdown has occurred. Enable file system journaling on Solaris 7 and 8Solaris 7 and 8 include a native implementation of file system journaling. This feature, known as "intent logging""logging" enables FASTER file system operation and FASTER system boot.Its trivial to implement and safe to use. The new logging feature is an option to the Unix File System (UFS), whichfile system for all disk partitions on SUN servers, except for partitions holding swap space. By default, the journalindisabled. Logging is enabled on a per file system basis, and it can even be enabled on / (root file system) and othersystem partitions.BackgroundSolaris UFS logging works by allocating space from the file systems free blocks. Within that space, all metadata chfile system are written. Metadata includes directory and I-node information but not file data blocks, essentially everactual data within the file. So, for example, a "file create" modifies the directory structure and allocates a new I-nodactivities are written to the logging space. Once the metadata changes are made to the logging area, the system is frother operations to the file system. In the background, the information in the log is flushed to the file system and upappropriate directory and I-node structures, completing the file system operations.The logging data is written sequentially within the log space. Its therefore much faster for the operating system to cmetadata changes via logging and background flushing than by directly modifying the metadata (via random I/O) spdisk. The size of the logging space is based on the size of the file system, and equals 1 MB per 1 GB of file systemMB. The space is used as a circular log: if the log space is about to fill up, new metadata change requests are pauseis emptied. As changes are moved from the log to the file system, that log space is made available, and new metadabe written to the logging space.Usually with UFS, if the system crashes during any file system operation, the entire system must have its consistenthe fsck command. That command can take several minutes per file system because it checks all metadata and file dthe structures are correct, free, and used, and that the I-node block counts are correct. It also confirms that the free sis current, repairs inconsistencies, and occasionally requires manual intervention to fix large problems. Files and evcan be lost, depending on the operations occurring at the time of the crash.Because metadata changes are made first to the log space rather than to the file system, the consistency check for asystem after a crash is a simple and fast operation. The system evaluates the logging data and determines which chacompleted against the underlying file system, which had yet to start, and which were in progress. Those completedstarted are removed from the log, and those partly completed are either undone or completed. If theres sufficient dacomplete the operation, its completed. Otherwise, the changes made are removed from the underlying file system.People familiar with database operation will recognize the similarity between database transaction processing and there. The end result is that the underlying file system is consistent, and no thorough consistency checking is neededoperation completes in a few seconds per file system.Using logging
  • Starting with Solaris 7, theres a new logging option to the mount command and in the /etc/vfstab system configuraLogging only appears in a couple other places within Solaris. The mount command shows which partitions are moulogging in the options fields for each partition on which logging is enabled. Finally, at system boot time, the fsck phpartition whether each is stable, logging, or being checked. There are no other status commands available to determlogging.A = Device to mountB = Device to fsckC = Mount pointD = Filesystem TypeE = Fsck pass (unimportatnt with logging)F = Mount at bootG = Mount options# ------------------------------------------------------------------#A B C D E F G# ------------------------------------------------------------------fd - /dev/fd fd - no -/proc - /proc proc - no -/dev/dsk/c0t0d0s3 - - swap - no -/dev/dsk/c0t0d0s0 /dev/rdsk/c0t0d0s0 / ufs 1 no logging/dev/dsk/c0t0d0s6 /dev/rdsk/c0t0d0s6 /usr ufs 2 no logging/dev/dsk/c0t0d0s1 /dev/rdsk/c0t0d0s1 /var ufs 3 no logging/dev/dsk/c0t0d0s7 /dev/rdsk/c0t0d0s7 /home ufs 4 yes logging/dev/dsk/c0t0d0s5 /dev/rdsk/c0t0d0s5 /opt ufs 5 yes logging/dev/dsk/c0t8d0s0 /dev/rdsk/c0t8d0s0 /u01 ufs 6 yes logging/dev/dsk/c0t9d0s0 /dev/rdsk/c0t9d0s0 /u02 ufs 7 yes logging/dev/dsk/c0t10d0s0 /dev/rdsk/c0t10d0s0 /u03 ufs 8 yes logging/dev/dsk/c0t11d0s0 /dev/rdsk/c0t11d0s0 /u04 ufs 9 yes logging/dev/dsk/c0t12d0s0 /dev/rdsk/c0t12d0s0 /u05 ufs 10 yes logging/dev/dsk/c1t13d0s0 /dev/rdsk/c1t13d0s0 /app ufs 11 yes logging/dev/dsk/c1t14d0s0 /dev/rdsk/c1t14d0s0 /users ufs 12 yes loggingswap - /tmp tmpfs - yes -Logging increases performance, decreases fsck time, removes the risk of a file system corruption, can be used on alpartitions (including root), and is free. Solaris Syslog Daemon DebuggingThe log system messages daemon syslogd reads and forwards system messages to the appropriate log files and/or uupon the priority of a message and the system facility from which it originates. The configuration file /etc/syslog.cowhere messages are forwarded. The syslogd daemon ignores any faulty entry in /etc/syslog.conf, specially spaces inare not recognized by syslogd. Therefore always check the entries in /etc/syslog.conf in the debugging mode of syslHow to check /etc/syslog.conf# /etc/init.d/syslog stop# /usr/sbin/syslogd -d
  • getnets() found 1 addresses, they are: 0.0.0.0.2.2amiloghost() testing 193.247.121.196.2.2cfline(*.err;kern.notice;auth.notice /dev/sysmsg)cfline(*.err;kern.debug;daemon.notice /var/adm/messages)cfline(mail.info;mail.debug /var/log/maillog)syslogd: line 14: unknown priority name "debug /var/log/maillog"cfline(*.alert;kern.err;daemon.err operator)cfline(*.alert root)cfline(*.emerg *)cfline(user.err /dev/sysmsg)cfline(user.err /var/adm/messages)cfline(user.alert root, operator)cfline(user.emerg *)syslogd: version 1.70Started: Sat Jan 6 10:11:47 2001Input message count: system 0, network 0# Outputs: 105 3 3 3 5 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 X CONSOLE: /dev/sysmsg7 3 3 5 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 X FILE: /var/adm/messagesX X 6 X X X X X X X X X X X X X X X X X X X X X X UNUSED:3 1 1 3 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 X USERS: operator1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 X USERS: root0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 X WALL:X 3 X X X X X X X X X X X X X X X X X X X X X X X CONSOLE: /dev/sysmsgX 3 X X X X X X X X X X X X X X X X X X X X X X X FILE: /var/adm/messagesX 1 X X X X X X X X X X X X X X X X X X X X X X X USERS: root, operatorX 0 X X X X X X X X X X X X X X X X X X X X X X X WALL: Per File StatisticsFile Tot Dups Nofwd Errs---- --- ---- ----- ----/dev/sysmsg 0 0 0 0/var/adm/messages 0 0 0 0 0 0 0 0operator 0 0 0 0root 0 0 0 0WALL 0 0 0 0/dev/sysmsg 0 0 0 0/var/adm/messages 0 0 0 0root,operator 0 0 0 0WALL 0 0 0 0syslogd: restartedoff & running....sys_poll blocking, init_cnt=0
  • # ^D# /etc/init.d/syslog startLine 14 in /etc/syslog.conf are filled up with spaces instead of tabs. Replace the spaces with tabs and syslogd will aentry in Line 14. Does each Oracle Process use more than 100M memoIf you check the oracle process with the OS comand "pmap" or "top", you can see that each oracle process use mormemory. Is this a problem on the Oracle installation or something else? It seems that pmap counts the SGA size asmemory segment of each oracle process, but we believe the SGA size should be shared.Output from "top" on our Solaris System with Orcale 8.1.7.0PID USERNAME THR PRI NICE SIZE RES STATE TIME CPU COMMAND-----------------------------------------------------------------361 oracle 258 59 0 124M 88M sleep 0:01 0.00% oracle373 oracle 11 59 0 122M 88M sleep 41:50 0.00% oracle363 oracle 11 59 0 119M 88M sleep 0:01 0.00% oracle365 oracle 11 58 0 119M 88M sleep 0:17 0.02% oracle359 oracle 1 59 0 119M 89M sleep 0:00 0.00% oracle377 oracle 1 59 0 119M 88M sleep 0:00 0.00% oracle375 oracle 1 58 0 119M 88M sleep 0:00 0.00% oracle367 oracle 1 58 0 118M 89M sleep 0:00 0.00% oracle371 oracle 1 58 0 118M 89M sleep 0:00 0.00% oracle369 oracle 1 58 0 118M 88M sleep 0:00 0.00% oracleMemory Allocation for Oracle ProcessesOn many UNIX platforms and specially on Sun platforms, the text of the Oracle binary and shared libraries are actubetween background processes if these instances share the same ORACLE_HOME. So you need to subtract the shaoracle binary and the shared libraries in the result of the OS commands.Even pmap and pmen utilities make mistakes between these memory divisions, and sometimes SGA and text execuadded incorrectly.Determine the memory used by each Oracle background process on a SolarisThis can be used by anyone who has privleges for the pmap, which can be found in /usr/proc/bin/. First, we need toprocess id (PID) of the Oracle background process you wish to determine the memory size for. This is done by issufollowing command:# ps -u oracle -f UID PID PPID C STIME TTY TIME CMDoracle 359 1 0 12:26:17 ? 0:00 ora_pmon_DIA3oracle 361 1 0 12:26:17 ? 0:01 ora_dbw0_DIA3oracle 363 1 0 12:26:17 ? 0:01 ora_lgwr_DIA3
  • oracle 365 1 0 12:26:17 ? 0:18 ora_ckpt_DIA3oracle 367 1 0 12:26:17 ? 0:01 ora_smon_DIA3oracle 369 1 0 12:26:17 ? 0:00 ora_reco_DIA3oracle 371 1 0 12:26:17 ? 0:00 ora_snp0_DIA3oracle 373 1 0 12:26:17 ? 41:50 ora_s000_DIA3oracle 375 1 0 12:26:17 ? 0:00 ora_d000_DIA3oracle 377 1 0 12:26:18 ? 0:00 ora_d001_DIA3Second, you then enter the following commands for the DB Writer process (ora_dbw0_DIA3) with process id = 36example.# /usr/proc/bin/pmap 361 | grep "shmid"80000000 82992K read/write/exec/shared [ shmid=0x2 ]# /usr/proc/bin/pmap 361 | grep "total"total 124232KThen you take the total size: 124232K and subtract the SGA size which the line marked with "shmid=" above, in th82992K. So, 124232K minus 82992K is 41240K. So, the DBWR background process is approximately 41.2 MB.steps for all the background processes. Sizing up Solaris Memory with the RMCmem PackagHow much memory is needed on SUN Solaris? Explaining memory in Solaris by reviewing the different types of mintroducing a set of tools, the RMCmem package.Install RMCmem PackageDownload the RMCmem tools available from ftp://playground.sun.com/pub/memtool. The package includes a kernprovides extra instrumentation.# cd /tmp# zcat RMCmem3.8.2.tar.gz | tar xvf -# pkgadd -d .The package is installed in /opt/RMCmem (see README in this directory)Virtual / Physical Memory UsageSolaris is a virtual memory system. The total amount of memory that you can use is increased by adding swap spacIf you ever see "out of memory" messages, adding swap space is the usual fix. Performance of the system is very dehow much physical memory (RAM) you have. If you dont have enough RAM to run your workload, performance drapidly.Physical memory usage can be classified into four groups: Kernel memory mapped into kernel address space Process memory is mapped into a process address space
  • Filesystem cache memory that is not mapped into any address space Free memory that is not mapped into any address spaceRMCmem includes a simple command to summarize this:# /opt/RMCmem/bin/prtmemTotal memory: 989 MegabytesKernel Memory: 60 MegabytesApplication: 110 MegabytesExecutable & libs: 42 MegabytesFile Cache: 757 MegabytesFree, file cache: 11 MegabytesFree, free: 6 MegabytesTotal physical memoryThe total physical memory can be seen using prtconf. Memory is allocated in units called pages, and you can use thcommand to see the size in bytes per page:# /usr/sbin/prtconf | grep MemoryMemory size: 1024 Megabytes# /usr/bin/pagesize8192Kernel memoryKernel memory is allocated to hold the initial kernel code at boot time, then grows dynamically as new device drivemodules are used. Kernel tables also grow dynamically, unlike some older versions of Unix. As you add hardware aa system, the kernel will grow. In particular, to keep track of all the memory in a system, the kernel allocates a pageIf you have several gigabytes of RAM this table gets quite large. The dynamic kernel memory allocator grabs mem"slabs," then allocates smaller blocks more efficiently. This means that the kernel tends to grab a bit more memoryusing. If there is a severe memory shortage, the kernel unloads unused kernel modules and devices and frees unusedsimplest summary of kernel memory usage comes from sar. To show the kernel memory allocation (KMA) activitiesar for more details).# sar -k 1SunOS diamond 5.7 Generic_106541-12 sun4u 04/28/01sml_mem alloc fail lg_mem alloc fail ovsz_alloc fail6873088 6044236 0 44818432 43761720 0 11231232 0Application process memoryApplication processes consist of an address space divided into segments, where each segment maps either to a file,memory (the swap space), System V shared memory, or a memory mapped device. The mapped files include the co
  • initialized data for the command and all its shared libraries.What we really want to know, is the amount of RAM used by each segment. This is shown by the pmem commandRMCmem package.# /opt/RMCmem/bin/pmem 361361: ora_dbw0_DIA3 Kbytes Resident Shared Private Permissions Mapped File 82992 82992 82992 - read/write/exec [shmid=0x2] 16 16 8 8 read/exec libc_psr.so.1 16 16 8 8 read/exec libmp.so.2 8 8 8 - read/write/exec libmp.so.2........ .. .. . ............... ........... 112 80 72 8 read/exec libelf.so.1 8 8 8 - read/write/exec libelf.so.1 16 16 8 8 read/exec libkvm.so.1 8 8 8 - read/write/exec libkvm.so.1-------- ------ ------ ------ ------ 124232 93040 92728 312Now we can see that the process address space size is 124232 kilobytes; 93040 kilobytes of that are currently residememory, wherein 92728 kilobytes are shared with other processes while 312 kilobytes are private. When this coonly the 312 kilobytes of private memory were taken from the free list.If we now go through all the processes on the system, add up how much private memory they use, and also add in tmemory for each mapped file, well know how much application memory is in use. This summary is shown by prtmin the beginning, and the detail is listed by the memps command in RMCmem.# /opt/RMCmem/bin/mempsPID Size Resident Shared Private Process... ....... ...... ...... .... .............359 118904k 93608k 92800k 808k ora_pmon_DIA3367 118184k 93152k 92704k 448k ora_smon_DIA3369 117928k 93120k 92704k 416k ora_reco_DIA3371 118040k 93136k 92720k 416k ora_snp0_DIA3365 119040k 93120k 92712k 408k ora_ckpt_DIA3377 118344k 93080k 92720k 360k ora_d001_DIA3363 119088k 93056k 92720k 336k ora_lgwr_DIA3375 118344k 93048k 92720k 328k ora_d000_DIA3361 124232k 93040k 92728k 312k ora_dbw0_DIA3373 121608k 93032k 92728k 304k ora_s000_DIA3Filesystem cache memoryThis is the part of memory that is most confusing, as it is invisible. You can only tell its there if you access the samit is quicker the second time.
  • The RMCmem package adds kernel instrumentation that counts up all the pages for each cached file. The memps -mlists the files that are cached in order of the amount of memory theyre consuming.One problem is that within the kernel, the file is only known by its inode number and filesystem mount point. The dpathname for the file may not be known.The RMCmem package tries to solve this problem by catching file names as files are opened (by interposing on thecode) and making an inode-to-name lookup cache in the kernel. This cache size is limited (to 8192 entries by defaumay have been opened before the kernel module was loaded, so it cant always find the name.# memps -m Size InUse E/F Filename21064k 21064k F /usr (inode 540488) 8184k 824k F /usr (inode 260922) 7752k 7752k F /usr (inode 540429) 7480k 7480k F /usr (inode 540428) 7480k 7480k F /usr (inode 540427) 6896k 6896k F /usr (inode 540450) .... .... . .... ...... ......... and so on down to lots of files ...# cd /usr# find . -inum 540488./local/jdbc/ora817/old/libserver8.aMore infos about the RMCmem package can be found here as PDF Using Sun Solaris Manuals directly from CD-ROMSolaris 8:cd /cdrom/sol_8_doc./ab2cd (Start)http://quorum:8888 (Using the Doc online)./ab2cd stop (Stop)Solaris 7:cd /cdrom/sol_7_1199_doc./ab2cd (Start)http://diamond:8888 (Using the Doc online)./ab2cd stop (Stop) DLT-TAPE UNIT INSTALLATION on Solaris 7/8/9
  • Installation Instructions will cover the installation of the DLT tape peripheral hardware and configuration of the syscommunicate with the DLT tape peripheral. In this example we use a «QUANTUM DLT7000».The Solaris system must have the appropriate SCSI interface for DLT drive to attached to, a SCSI single-ended DLattached only to a SCSI single-ended interface. The same is true for SCSI differential attachment. Solaris includes aefficiently communicate with SCSI tape drives, such as the DLT tape peripheral.Perform the installation as follows:Shut down your Sun workstation/server and power off the machine and all scsi-devices. Connect the DLT to the scsgood cables and make sure the bus is terminated correctly. Set the scsi-id; id 4 or 5 are the most common to use.If possible use a separate or underutilized SCSI bus for the DLT. Running the tape drive on the same bus as the disknever let you achieve any good throughput. You bought the DLT because of performance didnt you ?1. STOP-A (L1-A) Power on the devices/machine again and halt the boot process with (or press the BREAK key if you have an ASCII console).2. probe-scsi-all Verify that the drive is connected properly. Note: output from probe-scsi will not always be correct if you enter the PROM monitor by breaking the boot process!3. boot -rv Boot the system and log in as root. When booting you should see a message similar to these: "st1: ".4. cd /kernel/drv Change directory to /kernel/drv. Edit the st.conf file by adding the following: tape-config-list="QUANTUM DLT7000","Quantum DLT7000","DLT7-data"; DLT7-data = 1,0x38,0,0x8639,4,0x82,0x83,0x84,0x85,3;tape-config-list="<DLT tape unit>","<DLT reference name>","<DLTdata>" tape-config-list is a variable defined by a series of tape configuration parameters listed belo <DLT tape unit> is the vendor and product ID string for the DLT device. Depending on the DLT tape peripheral you are installing, you must insert the appropriate vendor an for <DLT tape unit> as described in the following table: DLT Tape Product <DLT tape unit> DLT7000 QUANTUM DLT7000 (Total string character count, including spaces, must equal 15). <DLT reference name> is a name you select that the system will use to identify the DLT devic reference does not change the DLT product ID. When the system boots, the reference name will be d
  • list of peripheral devices recognized by the system. <DLT-data> is a variable containing a series of additional DLT device configuration information. Yo name in place of the <DLT data> string. You will continue editing the st.conf file by defining the selected for <DLT data>. The definition depends on the DLT tape peripheral you are installing. F series unit add the following line: 1,0x38,0,0x8639,4,0x82,0x83,0x84,0x85,3;<DLT data> contains 10 parameters and are described following:1 The first parameter, is the version number and should not change.0x38 The second parameter, designates the DLT tape type as defined in /usr/include/sys/mtio.h. #define MT_ISOTHER 0x36 /* generic other type of tape drive */ #define MT_ISDLT 0x38 /* sun: SCSI DLT tape drive */0 The third parameter is the block size. Since the DLT tape drive uses variable block size, this value should be zero.0x8639 The fourth parameter, 0x8639, is a summation of values that represent selected device options. The table below lists the options and the corresponding value: Option Value ST_VARIABLE 0x0001 ST_BSF 0x0008 ST_BSR 0x0010 ST_LONG_ERASE 0x0020 ST_NOWS_EOD 0x0200 ST_NLOADABLE 0x0400 ST_NO_RECSIZE_LIMIT 0x8000 The man st page has more information about these and other possible device options. For certain applications, it may be necessary to consider adding or removing one or more of the device options.4 The fifth parameter, 4, defines the number of densities. The maximum definable number of densities is 4.0x82 The sixth, seventh, eighth and ninth parameter are used for system selection of tape densities. Use0x83 these values for a DLT 7000 Tape Drive.0x840x853 The tenth parameter defines which density the system will use as the default density. The sixth, seventh, eighth and ninth parameters in the <DLT data>string are referenced by the system as 0, 1, 2 and 3, respectively. The 3 value for the tenth parameter selects the 0x85 density code as the system default density.After editing the st.conf file, reboot the system:
  • 5. shutdown-i0-g0 Reboot the System boot -rv The -r switch in the boot command enables a kernel compile and includes creation of device special files used for communication with the DLT devi The -v switch enables verbose mode display of system bootup. With verbo mode, the system should indicate that the DLT tape peripheral is attached displaying the <DLT reference name>string you selected.6. mt -t /dev/rmt/0 status Enter the following command to verify the installation: Vendor TANDBERG Product DLT7000 tape drive: sense key(0x0)= No Additional Sense residual= 0 retries= 0 file no= 0 block no= 0The target drive designations assigned by Solaris may take on values higher than already established in the /dev/rThis is not a problem but during a boot -rv, Solaris does not remove tape device files for drives that are no longthe system. This can increase the effort in locating the device file for the configured drive, however, this can be mindeleting the tape device files:rm /dev/rmt/*then either boot the system with a:boot -rvor issue the following at the command line prompt:drvconfig -i st; tapesIf the DLTtape is the only drive on the system, its target assignment should be zero. The Solaris man pages have mon drvconfig and tapes. Reconfigure Devices on SolarisIf you remove or add a device on Solaris then the devices files must be recreated, either with boot -rv or devfsexample to renumber the logical tape drive devices do the following: Tape drives were numbered beginning with /dinstead of /dev/rmt/0. The physical devices pointed to by the logical /dev/rmt/[012] devices no longer existed, and wrenumber the valid devices beginning at /dev/rmt/0. 1. Cleanup non-existent tape drive devices with devfsadm. # devfsadm -C -c tape -v 2. Remove all /dev/rmt logical links. # rm -f /dev/rmt/* 3. Recreate all /dev/rmt logical links with devfsadm
  • # devfsadm -c tape -vdevfsadmdevfsadm(1M) maintains the /dev and /devices namespaces. It replaces the previous suite of devfs administration todrvconfig(1M), disks(1M), tapes(1M), ports(1M), audlinks(1M), and devlinks(1M).OPTIONSThe following options are supported:-C Cleanup mode. Prompt devfsadm to cleanup dangling /dev links that are not normally removed. If the -c option is also used, devfsadm only cleans up for the listed devices classes.-c device_class Restrict operations to devices of class device_class. Solaris defines the following value for device_class: disk, tape, port, audio, and pseudo. This option may be specified more than once to specify multiple device classes. OpenBoot DiagnosticsThe Solaris operating system gets the jumpstart for its booting from a hardware-level interface called the OpenBooOBP for short. OpenBoot at its heart has an interactive command interpreter with a varied set of functions. OBP iswhich is stored in the socketed startup PROM of the computer and consists of two parts, the PROM and the NVRAAs stated earlier while the PROM acts as the interface for access to diagnostics and drivers, the NVRAM consists ouser defined parameters. Non Volatile information like the system identification information, device aliases etc areNVRAM.The OpenBoot PROM is programmable and can be programmed based on Forth, which is an interactiveprogramming language much like shell scripting.The main tasks performed by the OpenBoot firmware are: Initializing and Testing system hardware ( POST , power on self test) Interactive Debugging Management of NVRAM Parameters Start the Operating System bootUseful commands at OK prompt.Dignostics : boot Generalbanner boot - boot kernel from default device. printenvthis command shows the following systems Factory default is to boot Display all variables and current values.hardware informatiion : Model, architecture, from DISK if present, otherwise from NET.processor,keyboard, openboot version, Serial setenv <variable>no. ethernet address & host id. boot net - boot kernel from network Set variable to the given value. boot cdrom - boot kernel from CD-ROMtest floppy - test floppy disk drive boot disk1:h - boot from disk1 partition h set-default <variable>test net - test network loopbacks boot tape - boot default file from tape Reset the value of variable to the factory defaulttest scsi - test scsi interface boot disk myunix -as - boot myunix fromtest-all test for all devices with selftest disk with flags "-as" set-defaultsmethod Reset variable values to the factory defaults.watch-clockShow ticks of real-time clock
  • watch-net DEVALIASMonitor network broadcast packetswatch-net-all ok>show-devsMonitor broadcast packets on all net interfaces ok cd /pci@1f,4000/scsi@3 ok .propertiesprobe-scsi ok lsShow attached SCSI devices f00809d8 tape f007ecdc diskprobe-scsi-all ok .speedShow attached SCSI devices for all host CPU Speed : 200.00MHzadapters- internal & external. UPA Speed : 100.00MHz PCI Bus A : 66Mhz PCI Bus B : 33MhzKey SequencesThese commands are disabled if the PROM security is on. Also, if your system has full security enabled, you cannot apply any of the suggestedcommands unless you have the password to get to the ok prompt.Stop – Bypass POST. This command does not depend on security-mode. (Note: some systems bypass POST as a default; in such cases, use Stop-D tstart POST.)Stop-A Abort.Stop-D - Enter diagnostic mode (set diag-switch? to true).Stop-F - Enter Forth on TTYA instead of probing. Use exit to continue with the initialization sequence. Useful if hardware is broken.Stop-N Reset NVRAM contents to default values.Start an OpenBoot Diagnostics<STOP A>OK setenv diag-switch? trueOK setenv auto-boot? falseOK reset-allOK test-all or obdiagConfigure Graphics Console (e.g. Sun XVR-100 Graphics Accelerator) instead of serial TTYAOK show-displaysSelect the graphics accelerator, e.g. bOK nvalias mydev <CTRL-Y>OK setenv output-device mydevOK setenv use-nvramrc? trueOK reset-all Why doesnt my .forward file work?OverviewIf you are having problems where you have created a $HOME/.forward file in your home directory to forward eaccount to another and it just wont forward them?Set correct PermissionsFirst make sure the file isnt group or world writable.
  • -rwxrwxr-x 1 zahn dba 0 Jan 9 12:17 .forward # wrong-rwxr-xr-x 1 zahn dba 0 Jan 9 12:17 .forward # OKLastly, make sure your home directory isnt group or world writable.drwxrwxr-x 14 zahn dba 4096 Jan 9 12:20 zahn # wrongdrwxr-xr-x 14 zahn dba 4096 Jan 9 12:20 zahn # OK Simple Shell Script to backup your FilesOverviewA backup strategy is more complex than creating a redundant copy of disk storage and considering the strategy a susuccessful backup strategy must detail how the backup media are rotated, how the media are archived, how the systrecovered, and what the backup software will do to create the backup. Although all parts of the backup strategy areimportant, this tip will focus on the backup script and will detail a flexible backup script that uses built-in Solaris sowhich create a reliable local backup of a Solaris machine.IntroductionThe backup script will accomplish the following goals: Create a backup archive that is as easy to restore a single file as it is to restore an entire file s The backup script will run autonomously. The only human intervention will be to swap medi output. The filesystems or directories to backup can be specified in the script. Using automounter yo specify remote filesystems. The script will create a detailed log of the backup. The script will send an abbreviated email summary of the backup to the administrator. After a successful backup, the script will verify to some extent the contents of the backup me The backup script will be able to run on any Solaris 2.6 or greater machine without modificaTools usedWe use the well known utilities TAR, GZIP and DD, because they are available on any Unix system. They are verand simple to use. In case of an emergency it is important to have a simple way to restore, independent of complexincompatible software releases.Magnetic Tape ControlThe utility MT sends commands to a tape drive. Many of these commands are familiar, but some are not. The script will usecommands. rewind – rewind the tape rewoffl – rewind the tape and eject it (go offline) eom – space to end of recorded media on tape weof – write count EOF marks at current position on tape status – display current status of tape
  • ScriptClick here for the Shell Script