Wednesday, December 21, 2011

Setup Oracle 11g for Solaris 10 - 64bit

setup vnc
svccfg -s application/x11/x11-server add display1
svccfg -s application/x11/x11-server:display1 addpg options application
svccfg -s application/x11/x11-server:display1 addpropvalue options/server astring: "/usr/X11/bin/Xvnc"
svccfg -s application/x11/x11-server:display1 addpropvalue options/server_args astring: '"SecurityTypes=None"'


mount iso for Solaris
root@TT4-BCCLML1-S # lofiadm -a /u01/sol-10-u10-ga2-sparc-dvd.iso

mkdir /dvd_iso
root@TT4-BCCLML1-S # mount -F hsfs -o ro /dev/lofi/1 /dvd_iso

root@TT4-BCCLML1-S # cd /dvd_iso/

Use umount command to unmount image:
# umount /mnt

Now remove/free block device:
# lofiadm -d /dev/lofi/1

/usr/sbin/prtconf | grep "Memory size"

/usr/sbin/swap -l

#check  64bit
/bin/isainfo -kv

uname -r

cat /etc/release

df -k /tmp

pkginfo -i SUNWarc SUNWbtool SUNWhea SUNWlibC SUNWlibms SUNWsprot \
  SUNWtoo SUNWi1of SUNWi1cs SUNWi15cs SUNWxwfnt

#cai goi
pkgadd -d /dvd_iso/Solaris_10/Product SUNWi1cs SUNWi15cs

nếu cài oracle trước đó
more /var/opt/oracle/oraInst.loc

root@kinhnghiem # more /var/opt/oracle/oraInst.loc

prtconf | grep Mem
Memory size: 32760 Megabytes
groupadd oinstall
groupadd dba
groupadd oper

useradd -g oinstall -G dba -d /export/home/oracle oracle
mkdir /export/home/oracle
chown oracle:oinstall /export/home/oracle
passwd -r files oracle

$ id -p
uid=59008(oracle) gid=10001(dba) projid=3(default)

By default, Oracle 10 will allocate 40% of the total system physical memory to create SGA and PGA. So for 32G system memory, the shmmax for Oracle 10 will be 0.4*32G = 12.8G.

40% 13105
50% 16442
75% 24536
100% 32764

# cat /etc/project
projmod -sK "project.max-shm-memory=(privileged,10G,deny)" oracle
projmod -p 100  -c 'lptrung add' -U oracle -G dba  -K 'project.max-shm-memory=(privileged,21G,deny)' oracle

projadd -p 100  -c 'lptrung add' -U oracle -G dba  -K 'project.max-shm-memory=(privileged,21G,deny)' projoracle

usermod -K project=projoracle oracle

projects -l

# su - oracle
$ id -p
uid=100(oracle) gid=100(dba) projid=100(group.dba)
$ prctl -n project.max-shm-memory -i process $$
process: 5754: -bash
NAME                    PRIVILEGE     VALUE     FLAG     ACTION    RECIPIENT
project.max-shm-memory  privileged    21.00GB     -       deny 
- Nếu cài 2 oracle 10g và 11g vào chung một server(nếu tổng ram server là 32G muốn cho oracle 10g: 10G ram, oracle 11g: 10G ram) thì  set shmsys:shminfo_shmmax là 20G ram

vi /et/system
set shmsys:shminfo_shmmax=21474836480
set semsys:seminfo_semmni=1024
set semsys:seminfo_semmsl=1024
set shmsys:shminfo_shmmni=100

Reboot lại hệ thống

mkdir -p /u01/app/oracle/product/
chown -R oracle:oinstall /u01

vi .profile

ulimit -s 32768
ulimit -n 4096
TMP=/tmp; export TMP
# Select the appropriate ORACLE_BASE
ORACLE_BASE=/u01/app/oracle; export ORACLE_BASE
ORACLE_SID=bcclml; export ORACLE_SID

the following command displays a summary of I/O activity ten times, at ten-second intervals:
sar -b 10 10

Fix error tcp_largest_anon_port ....

bash-3.00# ndd -set /dev/tcp tcp_largest_anon_port 65535
bash-3.00# ndd -set /dev/udp udp_smallest_anon_port 9000
bash-3.00# ndd -set /dev/udp udp_largest_anon_port 65500
bash-3.00# ndd -set /dev/tcp tcp_smallest_anon_port 9000
bash-3.00# ndd -set /dev/tcp tcp_largest_anon_port 65500

Tuesday, December 20, 2011

Map san xuống server solaris sử dụng UFS and ZFS

1. Sử dụng UFS
-- Kiểm tra ổ dĩa hiện tại
root@c4is-standby-s # echo |format

       0. c1t0d0
       1. c1t1d0
       2. c1t2d0
       3. c1t3d0
       4. c5t600A0B8000330D0A0000054748BF696Bd0
       5. c5t600A0B80004700CE0000057348BF6846d0
       6. c5t600A0B80004700CE0000057748BF6AD0d0
Specify disk (enter its number): Specify disk (enter its number):

-- Vào SAN tạo volumn sau đó map xuống host
--Sau khi map xuống xong kiểm tra ổ đĩa mới 

root@c4is-standby-s # echo |format

       0. c1t0d0
       1. c1t1d0
       2. c1t2d0
       3. c1t3d0
       4. c5t600A0B8000330D0A0000054748BF696Bd0
       5. c5t600A0B80004700CE00001D8C4EF0CD9Cd0
       6. c5t600A0B80004700CE0000057348BF6846d0
       7. c5t600A0B80004700CE0000057748BF6AD0d0
Specify disk (enter its number): Specify disk (enter its number):

-- Ổ đĩa hiện tại đc map xuống là
 5. c5t600A0B80004700CE00001D8C4EF0CD9Cd0

-- Xác định lại silde của partition cần format để chọn  newfs

root@c4is-standby-s # format
Searching for disks...done

c5t600A0B80004700CE00001D8C4EF0CD9Cd0: configured with capacity of 1024.00GB

       0. c1t0d0
       1. c1t1d0
       2. c1t2d0
       3. c1t3d0
       4. c5t600A0B8000330D0A0000054748BF696Bd0
       5. c5t600A0B80004700CE00001D8C4EF0CD9Cd0
       6. c5t600A0B80004700CE0000057348BF6846d0
       7. c5t600A0B80004700CE0000057748BF6AD0d0

--- chọn ổ đĩa mới thứ 5 mới gắn vào

Specify disk (enter its number): 5
selecting c5t600A0B80004700CE00001D8C4EF0CD9Cd0
[disk formatted]
Disk not labeled.  Label it now? yes

        disk       - select a disk
        type       - select (define) a disk type
        partition  - select (define) a partition table
        current    - describe the current disk
        format     - format and analyze the disk
        repair     - repair a defective sector
        label      - write label to the disk
        analyze    - surface analysis
        defect     - defect list management
        backup     - search for backup labels
        verify     - read and display labels
        inquiry    - show vendor, product and revision
        volname    - set 8-character volume name
        !     - execute , then return
format> p

        0      - change `0' partition
        1      - change `1' partition
        2      - change `2' partition
        3      - change `3' partition
        4      - change `4' partition
        5      - change `5' partition
        6      - change `6' partition
        select - select a predefined table
        modify - modify a predefined partition table
        name   - name the current table
        print  - display the current table
        label  - write partition map and label to the disk
        ! - execute , then return
partition> p
Current partition table (original):
Total disk sectors available: 2147467230 + 16384 (reserved sectors)

Part      Tag    Flag     First Sector          Size          Last Sector
  0       root    wm                34       128.00MB           262177   
  1       swap    wu            262178       128.00MB           524321   
  2 unassigned    wu                 0            0                0   
  3 unassigned    wm                 0            0                0   
  4 unassigned    wm                 0            0                0   
  5 unassigned    wm                 0            0                0   
  6        usr    wm            524322      1023.74GB           2147467229   
  8   reserved    wm        2147467230         8.00MB           2147483613   

partition> q

==> chú ý thấy vị trí thứ 6 dung lượng là 1G  => chính là s6

Tạo newfs

newfs /dev/rdsk/c5t600A0B80004700CE00001D8C4EF0CD9Cd0s6

-- Tạo một thư mục tên u04, sao đó mount vào
mount -F ufs /dev/dsk/c5t600A0B80004700CE00001D8C4EF0CD9Cd0s6 /u04

Chú ý: thay thế rdsk bằng dsk

Lệnh để clean cache device 

devfsadm -Cv

Thêm thông tin vào vfstab
vi /etc/vfstab

/dev/dsk/c5t600A0B80004700CE00001D8C4EF0CD9Cd0s6  /dev/rdsk/c5t600A0B80004700CE00001D8C4EF0CD9Cd0s6   /u04    ufs     2       yes     -

Mount và test thử vfstab

#umount /u04
#mount /u04

2. Sử dụng ZFS

1. Kiểm tra kết nối FC từ server đến SAN Switch

root@SMS-S # luxadm -e port
/devices/pci@7c0/pci@0/pci@1/pci@0,2/SUNW,emlxs@1/fp@0,0:devctl    CONNECTED
/devices/pci@7c0/pci@0/pci@1/pci@0,2/SUNW,emlxs@2/fp@0,0:devctl    CONNECTED

root@SMS-S # echo |format
Searching for disks...done

       0. c0t0d0
       1. c0t2d0
       2. c4t600A0B8000330D0A00001A964ED581DDd0
       3. c4t600A0B80004700CE00001CF54ED5754Cd0
Specify disk (enter its number): Specify disk (enter its number): 

Chú ý:  Nếu server chưa thấy Volume trên SAN thì dung lệnh devfsadm –Cv hoặc reboot server.
2. Khởi tạo volume với định dạng ZFS file

root@SMS-S # zpool create u03 c4t600A0B80004700CE00001CF54ED5754Cd0

invalid vdev specification
use '-f' to override the following errors:

zpool create -f diskbackup c6t600A0B800047006456456530057548BF69DEd0

root@SMS-S # zpool status -v u03
  pool: u03
 state: ONLINE
 scrub: none requested

        NAME                                     STATE     READ WRITE CKSUM
        u03                                      ONLINE       0     0     0
          c4t600A0B80004700CE00001CF54ED5754Cd0  ONLINE       0     0     0

errors: No known data errors
root@SMS-S #

3. Test tốc độ đọc ghi trên U03
Script test  u03

mkfile 10G testfile

root@SMS-S # date
Fri Jan 27 14:53:21 ICT 2012
root@SMS-S # date
Fri Jan 27 14:55:18 ICT 2012

# zpool iostat u03 1

u03         8.86G  1007G      0    551      0  68.6M
u03         8.86G  1007G      0    285      0  35.4M
u03         8.86G  1007G      0    327      0  40.7M
u03         8.86G  1007G      0    220      0  27.4M
u03         8.86G  1007G      0    380      0  47.1M
u03         8.86G  1007G      0    284      0  35.2M
u03         8.86G  1007G      0    413      0  51.5M
u03         8.86G  1007G      0  1.12K      0   142M
u03         8.86G  1007G      0  1.04K   128K   128M
u03         9.42G  1007G      0  1.03K      0   131M
u03         9.42G  1007G      0    831      0   103M
u03         9.42G  1007G      0    570      0  70.8M
u03         9.42G  1007G      0    594      0  73.6M
u03         9.42G  1007G      0    570      0  70.8M
u03         9.42G  1007G      0    725      0  90.0M
u03         9.99G  1006G      0    342      0  37.2M

Saturday, December 17, 2011

kloxo 500 - Internal Server Error

There are still have a problem when update Kloxo to version 6.1.10 (especially from 6.1.7) with using 'auto update' or 'update home' on kloxo.

Solution, run this commands:

# for sure to using latest version for certain packages
yum update
# shell command version for update
sh /script/upcp
# cleanup process for certain settings
sh /script/cleanup
# restart some services
service xinetd restart
# fix web config to make sure using latest
sh /script/fixweb --server=all
# Better reboot after that

Friday, December 16, 2011

Reset password root MYSQL với quyền root

[root@servert1 ~]# /etc/init.d/mysqld stop

Stopping MySQL: [ OK ]

[root@servert1 ~]# mysqld_safe --skip-grant-tables &

 [1] 13694

[root@servert1 ~]# Starting mysqld daemon with databases from /var/lib/mysql

[root@servert1 ~]# mysql -u root

Welcome to the MySQL monitor. Commands end with ; or \g.

Your MySQL connection id is 1

Server version: 5.0.77 Source distribution

Type 'help;' or '\h' for help. Type '\c' to clear the buffer.

mysql> use mysql;

mysql> update user set password=PASSWORD("testpass") where User='root';

Query OK, 3 rows affected (0.05 sec)

Rows matched: 3 Changed: 3 Warnings: 0

mysql> flush privileges;

Query OK, 0 rows affected (0.04 sec)

mysql> quit

Thursday, December 15, 2011

ORA-27102: out of memory - Set the kernel parameters

When trying to increase the SGA to approach half available RAM with an Oracle 64-bit version on a Linux 64-bit operating system, even though shmmax is set to match half the amount of  RAM, you get the following error when trying to start the instance:

SQL> startup nomount
ORA-27102: out of memory
Linux-x86_64 Error: 28: No space left on device  


shmall is too small, most likely is set to the default setting of 2097152
$ cat /proc/sys/kernel/shmall


shmall is the total amount of shared memory, in pages, that the system can use at one time.


Set shmall equal to the sum of all the SGAs on the system, divided by the page size.
The page size can be determined using the following command:
$ getconf PAGE_SIZE
For example, if the sum of all the SGAs on the system is 16Gb and the result of  '$ getconf PAGE_SIZE' is 4096 (4Kb) then set shmall to 4194304 pages
As the root user set the shmall to 4194304 in the /etc/sysctl.conf file:
kernel.shmall = 4194304

then run the following command:
$ sysctl -p
$ cat /proc/sys/kernel/shmall

The above command loads the new value and a reboot is not necessary.
Switch back to being the oracle user and retry the startup command.
Modifying /etc/sysctl.conf is a permanent workaround (applies at boot time). If for some reason you DO NOT want to change the system wide configuration, you can do it on the fly by directly changing the kernel pseudo FS AKA procfs.
e.g. echo "4194304" > /proc/sys/kernel/shmall

Using HUGEPAGES does not alter the calculation for configuring shmall.

 Set the kernel parameters
Add the following the lines in the file /etc/sysctl.conf
kernel.shmall = physical RAM size / pagesize For most systems, this will be the value 2097152. See Note 301830.1 for more information.
kernel.shmmax = 1/2 of physical RAM. This would be the value 2147483648 for a system with 4Gb of physical RAM.
kernel.shmmni = 4096
kernel.sem = 250 32000 100 128
fs.file-max = 512 x processes (for example 65536 for 128 processes)
Development recommends a minimum of 327679 for active systems.
net.ipv4.ip_local_port_range = 9000 65500
(The runInstaller (OUI) checks may expect this to be the old guidance of 1024 65000.
The new guidance from Oracle development is 9000 65500.
Please allow the runInstaller (OUI) to proceed with the new guidance from Oracle development.)
net.core.rmem_default = 262144
net.core.rmem_max = 2097152
net.core.wmem_default = 262144
net.core.wmem_max = 1048576
To place these changes into effect, execute the command
# sysctl -p

Wednesday, December 14, 2011

Linux Add a Swap File – Howto

Procedure To Add a Swap File

You need to use the dd command to create swap file. The mkswap command is used to set up a Linux swap area on a device or in a file.
a) Login as the root user.
b) Type following command to create 512MB swap file (1024 * 512MB = 524288 block size):
# dd if=/dev/zero of=/swapfile1 bs=1024 count=524288
c) Set up a Linux swap area:
# mkswap /swapfile1
d) Activate /swapfile1 swap space immediately:
# swapon /swapfile1
e) To activate /swapfile1 after Linux system reboot, add entry to /etc/fstab file. Open this file using a text editor such as vi:
# vi /etc/fstab
Append the following line:
/swapfile1 swap swap defaults 0 0
So next time Linux comes up after reboot, it enables the new swap file for you automatically.
g) How do I verify swap is activated or not?
Simply use the free command:
$ free -m

See also:

free command

Display free memory size in MB:
$ free -m

             total       used       free     shared    buffers     cached
Mem:           750        625        125          0         35        335
-/+ buffers/cache:        254        496
Swap:          956          0        956
Displays a line containing the totals memory in MB:

$ free -t -m

       total       used       free     shared    buffers     cached
Mem:           750        625        125          0         35        335
-/+ buffers/cache:        253        496
Swap:          956          0        956
Total:        1707        625       1082

vmstat command

Type vmstat command at shell prompt:
$ vmstat

procs -----------memory---------- ---swap-- -----io---- --system-- ----cpu----
r  b   swpd   free   buff  cache   si   so    bi    bo   in    cs us sy id wa
1  0      0 131620  35432 341496    0    0    42    82  737  1364 15  3 81  1

top command

Type top command at the shell prompt:

$ top
Sample outputs:


Tuesday, December 6, 2011

ORA-27102: out of memory on Solaris 10

This problem is due to insufficient shared memory from system while Oracle tries to create shared memory segment(depends on the size and SGA and PGA). Unlike earlier releases of Solaris (Solaris 8 and 9), most of the system parameters needed to run Oracle are already set properly, so the only one parameter to be set is the maximum shared memory. In earlier versions this was called shmsys:shminfo_shmmax and was set by editing the /etc/system file and rebooting. With Solaris 10 the parameter is set by modifying a "Resource Control Value". You can do this temporarily by using prctl, but that is lost at reboot so you will need to add the command to the oracle user. The other option is to create a default project for the oracle user:

$ projadd -U oracle -K "project.max-shm-memory=(priv,13G,deny)"

How large shared memory should be set for Oracle depends on the physical memory size. For Solaris , 8M is the default value on Solaris 9 and prior versions where as 1/4th of the physical memory is the default on Solaris 10 and later. Verifying the default setting of os is shown as the following
$ prtconf | grep Mem
Memory size: 32760 Megabytes

$ id -p
uid=59008(oracle) gid=10001(dba) projid=3(default)

$ prctl -n project.max-shm-memory -i project 3
project: 3: default
NAME    PRIVILEGE       VALUE    FLAG   ACTION                       RECIPIENT
        privileged      7.84GB      -   deny                                 -
        system          16.0EB    max   deny                                 -

By default, Oracle 10 will allocate 40% of the total system physical memory to create SGA and PGA. So for 32G system memory, the shmmax for Oracle 10 will be 0.4*32G = 12.8G.
(Temporary method)
$ prctl -n project.max-shm-memory -r -v 10G -i project 3

(No need to reboot)
$ projadd -p 100  -c 'test shmmax' -U oracle -G dba  -K 'project.max-shm-memory=(privileged,13G,deny)'

$ projects -l
        projid : 100
        comment: "test shmmax"
        users  : oracle
        groups : dba
        attribs: project.max-shm-memory=(privileged,13958643712,deny)

$ cat /etc/project shmmax:oracle:dba:project.max-shm-memory=(privileged,13958643712,deny)

- Nếu cài 2 oracle 10g và 11g vào chung một server(nếu tổng ram server là 32G muốn cho oracle 10g: 10G ram, oracle 11g: 10G ram) thì  set shmsys:shminfo_shmmax là 20G ram

vi /et/system
set shmsys:shminfo_shmmax=21474836480
set semsys:seminfo_semmni=1024
set semsys:seminfo_semmsl=1024
set shmsys:shminfo_shmmni=100

Reboot lại hệ thống

Sunday, December 4, 2011

Command Solaris


Akadia Information Technology

SUN Free Software

Unter findet man "ready to use" Software für SUN Solaris, wie beispielsweise TOP, AMANDA, GCC, GDB etc. Download via FTP von: «»

PCNFS installieren

CD-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
  • 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 Kabel
OK boot cdrom - w (Terminal an ttya)

How to Backup a System

$ init 0
OK boot -s
$ fsck -m /dev/dsk/c0t0d0s0 (und übrige Filesysteme)
$ tar cvf /dev/rmt/0l

Packages (Software die installiert wurde)

Anzeige der installierten Packages
$ pkginfo
Check ob Package SUNWpcnfd correct installiert ist
$ pkgchk -v SUNWpcnfd
Package installieren (Path ist meistens /cdrom/cdrom0/....)
$ pkgadd -d SUNWpcnfd
Das Package SUNWpcnfd entfernen
$ pkgrm SUNWpcnfd

Monitor Mode (OK Prompt)

In single user mode booten
OK boot -s
Kernel zwingen /devices neu aufzubauen nach dem Anschluss von neuer Hardware
OK boot -r
Detaillierter Bootvorgang
OK boot -v
Vom CD-ROM aus booten: Notboot !
OK boot cdrom
Angeschlossene SCSI-Geräte testen
OK probe-scsi
List System Devices, e.g. SUNW,hme = Sun Fast Ethernet PCI Adapter
OK show-devs
List Network Devices
OK show-nets
Monitoring Network Activity
OK apply watch-net
Monitor Variablen ändern, anzeigen
OK eeprom
OK eeprom ttya-mode=38400,8,n,1,h

Kernel Analyse

Welche Kernel-Module sind geladen ?
$ modinfo
Kernel Konfiguration
Logfile von syslog
Konfiguration des syslog Daemon

Defaults einstellen

Directory mit Default files
Remote root logins erlauben
Timezone setzen

Wichtige Konfigurationsfiles

Master-File beim Booten
Run-Level Start/Stop Files
Scripts für Run-Levels

Admin Kommandos

$ shutdown -g0 -i0
$ reboot (entspricht init 6)
Konfiguration der LAN-Interfaces
$ ifconfig -a
  • Netmask setzen: siehe /etc/netmasks
  • Jedes LAN-Interface hat /etc/hostname.le0 mit Hostnamen
  • LAN-Setup: /etc/rcS.d/ (Interfaces konfigurieren)
    /etc/rc2.d/S72inetsvc (LAN konfigurieren)
Phys Addressen nachschauen
$ arp -a
Net to Media Table

Device IP Address           Mask            Flags Phys Addr
------ -------------------- --------------- ----- ---------------
le0    rabbit            00:60:08:57:17:86
le0    quorum      SP    08:00:20:89:27:03
le0    arkum             00:a0:24:4b:60:1c

IP-Routing konfigurieren

Alle hosts im Netz werden über den Router erreicht.
$ route add net 1
Der host wird über den Router erreicht. Siehe File /etc/rc2.d/S79staticroutes.
$ route add host 1
Routing Tabelle kontrollieren
$ netstat -nr

DNS konfigurieren

Angabe des DNS Nameservers
Reihenfolge definieren

Anonymous FTP aufsetzen

Siehe 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 /software
NFS-Server wird in /etc/init.d/nfs.client start gestartet.
Anzeige welche Directories gondwana zum mounten freigegeben hat
dfshares gondwana
gondwana:/export/home/zahn gondwana - -
gondwana:/export/home/steiner gondwana - -

NFS-Server Konfiguration

/etc/dfs/dfstab (Soll) --> /etc/dfs/sharetab
Directory read-only freigeben
$ share -o ro /usr/software
Alle Directories in /etc/dfs/dfstab freigeben
$ shareall
Alle Directories in /etc/dfs/dfstab zurücknehmen
$ unshareall
NFS-Server wird gestartet in
Anzeige der freigegbenen lokalen Direcories
$ share
Anzeige welche Clients nutzen welche Directories eines NFS-Servers
$ dfmounts -F nfs gondwana

gondwana /export/home/zahn,


- /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 anschliessen

Beispiel: 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 hinten am Gerät ein Tippschalter eingestellt werden. Man beacht, dass in der Regel das letzte Gerät terminiert werden muss.
  • Disk formatieren (nur wenn notwendig !)

    In der Regel muss eine Disk nicht neu formatiert werden, ist dies jedoch notwendig so steht unter Solaris das Utility format zur Verfügung.


0. c0t0d0
1. c0t1d0
2. c0t2d0
3. c0t4d0
4. c0t5d0

Specify disk (enter its number): 4

format> type

0. Auto configure
1. DDRS-39130
2. Seagate ST118273
3. Seagate ST150176L
4. Quantum ProDrive 80S
5. Quantum ProDrive 105S
6. CDC Wren IV 94171-344
7. SUN0104
8. SUN0207
9. SUN0327
10. SUN0340
11. SUN0424
12. SUN0535
13. SUN0669
14. SUN1.0G
15. SUN1.05
16. SUN1.3G
17. SUN2.1G
18. SUN2.9G
19. IBM-DDRS-39130-S71D
20. SEAGATE-ST118273N-5764
21. SEAGATE-ST150176LW-0002
22. other
Specify disk type (enter its number)[21]: 21

format> format (confirm with "yes")
  • Disk partitionieren
Dadurch wird die Disk in logische Teile unterteilt. Jeder teil enthält ein eigenes Filesystem.
format> part
Nun die Partitionierungsdaten eingeben, zB
partition> print

Current partition table (original):
Total disk cylinders available: 2733 + 2 (reserved cylinders)

Part Tag Flag Cylinders Size Blocks
0 root wm 0 - 204 152.15MB (205/0/0) 311600
1 swap wu 205 - 377 128.40MB (173/0/0) 262960
2 backup wm 0 - 2732 1.98GB (2733/0/0) 4154160
3 home wm 378 - 1017 475.00MB (640/0/0) 972800
4 unassigned wm 0 0 (0/0/0) 0
5 unassigned wm 1018 - 1928 676.13MB (911/0/0) 1384720
6 usr wm 1929 - 2732 596.72MB (804/0/0) 1222080
7 unassigned wm 0 0 (0/0/0) 0
  • Label erzeugen (aktuelle Partitionierung speichern)
partition> label
partition> y
partition> quit
format> quit
  • Filesystem erstellen
newfs -v -m 0 /dev/rdsk/c0t5d0s0
Damit wird ein Filesystem mit 0 % Min-Free auf der Partition 0 der Disk an der SCSI-Adresse 5 erstellt.
  • Filesystem mounten
Dazu 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 installed

Multiply the Blocks column by 512
$ swap -l
swapfile            dev swaplo blocks    free
/dev/dsk/c0t0d0s1 32,1     16 262944 262944

262944 * 512 = 134 MB

Show Operating System Patch Level

$ showrev -p

Patch: 105181-16
Note, that Patchlevel 105181-15 is minimal needed for Oracle 8.1.6

How to install a Sun Solaris Jumbo Patch ?

- Download the Patch from:
- Read the README File included in the Patch
- Usually the only thing you have to do is:
$ cd
$ ./install_custer $ cat /var/sadm/install_data/_log $ showrev -p
Reboot the system

Tracing System Calls

You can trace system calls with truss on Solaris an strace on Linux
$ truss svrmgrl

Troubleshooting Solaris Device Files

If you suspect troubles with your Solaris device files, e.g. system doesn't boot after a filesystem check, you may repair the solaris system using the following commands.
Halt the system immediately with the keys STOP-A, you will now see the boot prompt: OK
Reset the machine with
OK reset
Boot the machine with
OK boot -r
The command boot -r will rebuild all devices files according to your attached hardware. If you cannot boot the machine, you can try the following commands: drvconfig, disks, tapes
drvconfig - configure the /devices directory
The default operation of drvconfig is to create the /devices directory tree that describes, in the filesystem namespace, the hardware layout of a particular machine. Hardware devices present on the machine and powered on as well as pseudo-drivers are represented under /devices. Normally this command is run automatically after a new driver has been installed (with add_drv(1M)) and the system has been rebooted.
disks - creates /dev entries for hard disks attached to the system
Disks creates symbolic links in the /dev/dsk and /dev/rdsk directories pointing to the actual disk device special files under the /devices directory tree.
tapes - creates /dev entries for tape drives attached to the system

Tapes creates symbolic links in the /dev/rmt directory to the actual tape device special files under the /devices directory tree. Tapes searches the kernel device tree to see what tape devices are attached to the system.

Short Tips to maintain Sun Solaris

Here are some short tips for common tasks on SUN Solaris 2.6, 7 and 8
Important 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 mode
Show currently mounted filesystems
# /etc/mnttab: Contains information about devices that
# are currently mounted. If there are mounted filesystems
# with quotas enabled, display them
if /usr/bin/cut -f 4 /etc/mnttab | \
  /usr/bin/egrep '^quota|,quota' >/dev/null 2>&1; then
  echo 'There are mounted filesystems with quotas enabled'
How 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 are

cat /etc/nodename
cat /etc/hostname.*
cat /etc/default/init
cat /etc/defaultdomain
cat /etc/inet/hosts
cat /etc/inet/netmasks
How to configure Asynchronous PPP ?
Configure /etc/ for the aspppd daemon
$ /usr/sbin/aspppd -d 1
How 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 0
Enable IP Forwarding (Machine acting as a Router)
$ /usr/sbin/ndd -set /dev/ip ip_forwarding 1
How 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

# Show Default Route

$ /usr/sbin/route -fn
default done
How 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`"
RPC (Remote Procedure Call) Configuration
# rpcbind - universal addresses to RPC program number mapper
# rpcinfo - report RPC information
Solaris 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/keyserv
How 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 &
Where to find syslogd messages ?
Configuration File: /etc/syslog.conf
Message File:       /var/adm/messages

IP-Aliasing for SUN Solaris

# How to setup IP-Alias on SUN Solaris
1. Setup File /etc/hostname.hme0:1 for the second IP-Address
   cat /etc/hostname.hme0:1
2. Insert IP-Address in /etc/hosts
   # Internet host table
   #      ux-portal1       # IP-address on hme0:0      ldap             # IP-alias on hme0:1
3. Start alias IP-Address on Interface in /etc/rc2.d
   S99ipalias -> ../init.d/ipalias
   # 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 ]
     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 up
        'stop')  # Stop second IP address on hme0:1
                 echo "Stop multi-homed server for UX-ALIAS1 on hme0:1"
                 ifconfig hme0:1 down
4. Check IP-Address on second Interface
   ifconfig -a

Solaris automounter installs filesystems by default in /net

The Solaris automount utility installs autofs mount points and associates an automount map with each mount point. The autofs file system monitors attempts to access directories within it and notifies the automountd daemon. The daemon uses the map to locate a file system, which it then mounts at the point of reference within the autofs file system. You can assign a map to an autofs mount using an entry in the /etc/auto_master map or a direct map in /etc/auto_direct. If the file system is not accessed within an appropriate interval (five minutes by default), the automountd daemon unmounts the file system.
Default Mapping under /net
The mount point /net is by default the location, where automountd mounts NFS filesystems, which are exported on other machines. Lets suppose, that you have the filesystem /home exported on the NFS server saphir, then the (Solaris) NFS client with an active automounter will automatically mount this NFS filesystem under /net/saphir/.
Mapping using /etc/auto_direct
You probably doesn't 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 under /opt/local.

Solaris keyboard utility

The Solaris utility kbd manipulates the state of the keyboard or display the type of keyboard or change the default keyboard abort sequence effect. Suppose, that you do not want that everybody can halt the the system you must change the default value. We also noticed, that the Solaris machines attached to a switch box, using a character terminal on a serial line, may halt when you switch from one machine to the other.
kbd [ -r ] [ -t ] [ -c on|off ]
    [ -a enable|disable ] [ -d keyboard device ]
kbd -i [ -d keyboard device ]
kbd manipulates the state of the keyboard, or displays the keyboard type or allows the default keyboard abort sequence effect to be 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 default file, /etc/default/kbd. Only keyboards that support a clicker respond to the -c option. If you want to turn clicking on by default, add or change the current value of the KEYCLICK variable to the value on in the keyboard default file, /etc/default/kbd, as shown here.
Then, run the command 'kbd -i' to change the current setting. Valid settings for this variable are the values on and off. Other values 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 on most systems) 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 and enter the debugger or the monitor. 
If you want to permanently change the software default effect of the keyboard abort sequence, you can add or change the current value of the KEYBOARD_ABORT variable to the
value disable in the keyboard default file, /etc/default/kbd, as shown here.
Then, run the command 'kbd -i' to change the current setting. Valid settings for this value are the values enable and disable. Other values are ignored. If the variable is not specified in the default file, the setting is unchanged.

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 and process keyclick and keyboard abort default values from the /etc/default/kbd file. This option can only be used by the superuser.

Reset the keyboard as if power-up.
Return the type of the keyboard being used.
On/Off state Turn the clicking of the keyboard on or off.
Enable/Disable state; Enable or disable the keyboard abort sequence effect.

Monitoring Performance

This chapter describes procedures for monitoring system performance by using the vmstat, iostat, df, and sar commands. 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     cpu
r b w  swap free re  mf  pi  po  fr de sr f0 s3 -- --  in  sy  cs us sy  id
0 0 8 28312  668  0   9   2   0   1  0  0  0  1  0  0  10  61  82  1  2  97
0 0 3 31940  248  0  10  20   0  26  0 27  0  4  0  0  53 189 191  6  6  88
0 0 3 32080  288  3  19  49   6  26  0 15  0  9  0  0  75 415 277  6 15  79
0 0 3 32080  256  0  26  20   6  21  0 12  1  6  0  0 163 110 138  1  3  96
0 1 3 32060  256  3  45  52  28  61  0 27  5 12  0  0 195 191 223  7 11  82
0 0 3 32056  260  0   1   0   0   0  0  0  0  0  0  0   4  52  84  0  1  99
Field Name
Reports the following states:

The number of kernel threads in the dispatch queue

Blocked kernel threads waiting for resources

Swapped out LWPs waiting for processing resources to finish

Reports on usage of real and virtual memory:

Available swap space

Size of the free list

Reports on page faults and paging activity, in units per second:

Pages reclaimed

Minor and major faults

Kbytes paged in

Kbytes paged out

Kbytes freed

Anticipated memory needed by recently swapped-in processes

Pages scanned by page daemon (not currently in use). If sr does not equal zero, the page daemon has been running.

Reports the number of disk operations per second, showing data on up to four disks

Reports the trap/interrupt rates (per second):

Interrupts per second

System calls per second

CPU context switch rate

Reports on the use of CPU time:

User time

System time

Idle time
How to Display System Event Information
Run 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 out
409376480 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) faults
409376480 minor (as) faults
  2957386 major faults
102738273 copy-on-write faults
 61711047 zero fill page faults
1002562077 pages examined by the clock daemon
     7881 revolutions of the clock hand
 16716370 pages freed by the clock daemon
  4999048 forks
  1138206 vforks
  5747009 execs
741660225 cpu context switches
736047593 device interrupts
528054538 traps
2496638575 system calls
430283487 total name lookups (cache hits 95%)
    81727 toolong
 10484677 user   cpu
  9528364 system cpu
443762786 idle   cpu
 16281790 wait   cpu
How to Display Swapping Statistics
Run vmstat -S to show swapping statistics.
 procs     memory            page            disk          faults      cpu
 r b w   swap  free  si  so pi po fr de sr m1 m3 m4 m5   in   sy   cs us sy id
 0 0 0   8512   888   0   0 12 21 55  0 417 1  0  0  0  206 1040  308  2  2 96
si = Average number of LWPs swapped in per second
so = Number of whole processes swapped out
How to Display Disk Utilization Information (iostat)
You can display disk activity information by using the iostat command with a time interval. The following example shows disk statistics gathered every five seconds.
iostat 5
      tty          md1           md3           md4           md5          cpu
 tin 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 98
For Each ...
Field Name

Number of characters in the terminal input queue

Number of characters in the terminal output queue

Blocks per second

Transactions per second

Average service time, in milliseconds

In user mode

In system mode

Waiting for I/O

How to Display Extended Disk Statistics
Run iostat -xtc to get extended disk statistics. This command displays a line of output for each disk.
                               extended device statistics      tty         cpu
device    r/s  w/s   kr/s   kw/s wait actv  svc_t  %w  %b  tin tout us sy wt id
md1       0.4  0.9    3.6    6.9  0.0  0.0   27.7   1   1    0    2  2  2  3 92
md3       0.1  0.2    1.0    1.3  0.0  0.0   21.7   0   0
md4       0.0  0.0    0.0    0.0  0.0  0.0    0.0   0   0
md5       0.0  0.0    0.7    0.0  0.0  0.0    9.9   0   0
md8       0.8  0.3    6.7   14.2  0.0  0.0   13.1   0   1
md10      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   1
md30      0.0  0.2    0.5    1.3  0.0  0.0   11.4   0   0
md31      0.0  0.2    0.5    1.3  0.0  0.0   10.2   0   0
md40      0.0  0.0    0.0    0.0  0.0  0.0    0.0   0   0
md41      0.0  0.0    0.0    0.0  0.0  0.0    0.0   0   0
md50      0.0  0.0    0.4    0.0  0.0  0.0    9.4   0   0
md51      0.0  0.0    0.4    0.0  0.0  0.0    7.3   0   0
md80      0.4  0.3    3.3   14.2  0.0  0.0   10.3   0   0
md81      0.4  0.3    3.3   14.2  0.0  0.0   11.7   0   1
sd0       0.6  2.1    6.0   22.8  0.0  0.0   16.3   0   3
sd1       0.6  2.1    6.0   22.8  0.0  0.0   15.2   0   2
Field Name
Reads per second
Writes per second
Kbytes read per second
Kbytes written per second
Average number of transactions waiting for service (queue length)
Average number of transactions actively being serviced
Average service time, in milliseconds
Percentage of time the queue is not empty
Percentage of time the disk is busy
How to Check CPU Utilization (sar)
Display CPU utilization with the sar -u command. (The sar command without any options is equivalent to sar -u.) At any given moment, the processor is either busy or idle. When busy, the processor is in either user or system mode. When idle, the processor 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 1
Measure CPU utilization during 60 secs 1440 times and write result in file sar.log.
sar -u -o sar.log 60 1440 
To later review disk and tape activity from that period:
sar -d -f sar.log
Field Name
Lists the percentage of time that the processor is in system mode
Lists the percentage of time that the processor is in user mode
Lists the percentage of time the processor is idle and waiting for I/O completion
Lists the percentage of time the processor is idle and is not waiting for I/O
A high %wio generally means a disk slowdown has occurred.

  Enable file system journaling on Solaris 7 and 8

Solaris 7 and 8 include a native implementation of file system journaling. This feature, known as "intent logging" or just "logging" enables FASTER file system operation and FASTER system boot. 
It's trivial to implement and safe to use. The new logging feature is an option to the Unix File System (UFS), which is the standard file system for all disk partitions on SUN servers, except for partitions holding swap space. By default, the journaling option is disabled. Logging is enabled on a per file system basis, and it can even be enabled on / (root file system) and other operating system partitions.
Solaris UFS logging works by allocating space from the file system's free blocks. Within that space, all metadata changes to the file system are written. Metadata includes directory and I-node information but not file data blocks, essentially everything but the actual data within the file. So, for example, a "file create" modifies the directory structure and allocates a new I-node, and those activities are written to the logging space. Once the metadata changes are made to the logging area, the system is free to perform other operations to the file system. In the background, the information in the log is flushed to the file system and updates the appropriate directory and I-node structures, completing the file system operations.
The logging data is written sequentially within the log space. It's therefore much faster for the operating system to complete metadata changes via logging and background flushing than by directly modifying the metadata (via random I/O) spread across the disk. The size of the logging space is based on the size of the file system, and equals 1 MB per 1 GB of file system space, up to 64 MB. The space is used as a circular log: if the log space is about to fill up, new metadata change requests are paused while the log is emptied. As changes are moved from the log to the file system, that log space is made available, and new metadata changes can be written to the logging space.
Usually with UFS, if the system crashes during any file system operation, the entire system must have its consistency checked via the fsck command. That command can take several minutes per file system because it checks all metadata and file data to ensure the structures are correct, free, and used, and that the I-node block counts are correct. It also confirms that the free space available is current, repairs inconsistencies, and occasionally requires manual intervention to fix large problems. Files and even directories can 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 a logged file system after a crash is a simple and fast operation. The system evaluates the logging data and determines which changes had completed against the underlying file system, which had yet to start, and which were in progress. Those completed or not yet started are removed from the log, and those partly completed are either undone or completed. If there's sufficient data in the log to complete the operation, it's 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 the activities here. The end result is that the underlying file system is consistent, and no thorough consistency checking is needed. That operation completes in a few seconds per file system.
Using logging
Starting with Solaris 7, there's a new logging option to the mount command and in the /etc/vfstab system configuration file. Logging only appears in a couple other places within Solaris. The mount command shows which partitions are mounted and lists logging in the options fields for each partition on which logging is enabled. Finally, at system boot time, the fsck phase reports per partition whether each is stable, logging, or being checked. There are no other status commands available to determine the state of logging.
A = Device to mount
B = Device to fsck
C = Mount point
D = Filesystem Type
E = Fsck pass (unimportatnt with logging)
F = Mount at boot
G = 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 logging
swap                 -                 /tmp    tmpfs  -  yes -
Logging increases performance, decreases fsck time, removes the risk of a file system corruption, can be used on all UFS partitions (including root), and is free.

Solaris Syslog Daemon Debugging

The log system messages daemon syslogd reads and forwards system messages to the appropriate log files and/or users, depending upon the priority of a message and the system facility from which it originates. The configuration file /etc/syslog.conf controls where messages are forwarded. The syslogd daemon ignores any faulty entry in /etc/syslog.conf, specially spaces instead of tabs are not recognized by syslogd. Therefore always check the entries in /etc/syslog.conf in the debugging mode of syslogd.
How to check /etc/syslog.conf
# /etc/init.d/syslog stop # /usr/sbin/syslogd -d
getnets() found 1 addresses, they are:
amiloghost() testing
cfline(*.err;kern.notice;auth.notice      /dev/sysmsg)
cfline(*.err;kern.debug;daemon.notice     /var/adm/messages)
cfline(;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 1.70
Started: Sat Jan  6 10:11:47 2001
Input message count: system 0, network 0
# Outputs: 10
5 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/sysmsg
7 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/messages
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: operator
1 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: root
0 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/sysmsg
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 FILE: /var/adm/messages
X 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, operator
                Per File Statistics
File                            Tot     Dups    Nofwd   Errs
----                            ---     ----    -----   ----
/dev/sysmsg                     0       0       0       0
/var/adm/messages               0       0       0       0
        0       0       0       0
operator                        0       0       0       0
root                            0       0       0       0
WALL                            0       0       0       0
/dev/sysmsg                     0       0       0       0
/var/adm/messages               0       0       0       0
root,operator                   0       0       0       0
WALL                            0       0       0       0
syslogd: restarted
off & running....
sys_poll blocking, init_cnt=0
# ^D
# /etc/init.d/syslog start
Line 14 in /etc/syslog.conf are filled up with spaces instead of tabs. Replace the spaces with tabs and syslogd will accept the new entry in Line 14.

Does each Oracle Process use more than 100M memory ?

If you check the oracle process with the OS comand "pmap" or "top", you can see that each oracle process use more than 100M memory. Is this a problem on the Oracle installation or something else? It seems that pmap counts the SGA size as the private memory segment of each oracle process, but we believe the SGA size should be shared.
Output from "top" on our Solaris System with Orcale
361 oracle   258  59    0  124M   88M sleep   0:01  0.00% oracle
373 oracle    11  59    0  122M   88M sleep  41:50  0.00% oracle
363 oracle    11  59    0  119M   88M sleep   0:01  0.00% oracle
365 oracle    11  58    0  119M   88M sleep   0:17  0.02% oracle
359 oracle     1  59    0  119M   89M sleep   0:00  0.00% oracle
377 oracle     1  59    0  119M   88M sleep   0:00  0.00% oracle
375 oracle     1  58    0  119M   88M sleep   0:00  0.00% oracle
367 oracle     1  58    0  118M   89M sleep   0:00  0.00% oracle
371 oracle     1  58    0  118M   89M sleep   0:00  0.00% oracle
369 oracle     1  58    0  118M   88M sleep   0:00  0.00% oracle
Memory Allocation for Oracle Processes
On many UNIX platforms and specially on Sun platforms, the text of the Oracle binary and shared libraries are actually shared between background processes if these instances share the same ORACLE_HOME. So you need to subtract the shared text of the oracle 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 executable are often added incorrectly.
Determine the memory used by each Oracle background process on a Solaris
This can be used by anyone who has privleges for the pmap, which can be found in /usr/proc/bin/. First, we need to find the process id (PID) of the Oracle background  process you wish to determine the memory size for. This is done by issueing the following command:
# ps -u oracle -f
oracle   359     1  0 12:26:17 ?        0:00 ora_pmon_DIA3
oracle   361     1  0 12:26:17 ?        0:01 ora_dbw0_DIA3
oracle   363     1  0 12:26:17 ?        0:01 ora_lgwr_DIA3
oracle   365     1  0 12:26:17 ?        0:18 ora_ckpt_DIA3
oracle   367     1  0 12:26:17 ?        0:01 ora_smon_DIA3
oracle   369     1  0 12:26:17 ?        0:00 ora_reco_DIA3
oracle   371     1  0 12:26:17 ?        0:00 ora_snp0_DIA3
oracle   373     1  0 12:26:17 ?       41:50 ora_s000_DIA3
oracle   375     1  0 12:26:17 ?        0:00 ora_d000_DIA3
oracle   377     1  0 12:26:18 ?        0:00 ora_d001_DIA3
Second, you then enter the following commands for the DB Writer process (ora_dbw0_DIA3) with process id = 361 as an example.
# /usr/proc/bin/pmap 361 | grep "shmid" 80000000 82992K read/write/exec/shared [ shmid=0x2 ]
# /usr/proc/bin/pmap 361 | grep "total"
total 124232K
Then you take the total size: 124232K and subtract the SGA size which the line marked with "shmid=" above, in this case it is 82992K. So, 124232K minus 82992K is 41240K. So, the DBWR background process is approximately 41.2 MB. Repeat this steps for all the background processes.

Sizing up Solaris Memory with the RMCmem Package

How much memory is needed on SUN Solaris? Explaining memory in Solaris by reviewing the different types of memory and introducing a set of tools, the RMCmem package.
Install RMCmem Package
Download the RMCmem tools available from The package includes a kernel module that provides 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 Usage
Solaris is a virtual memory system. The total amount of memory that you can use is increased by adding swap space to the system. If you ever see "out of memory" messages, adding swap space is the usual fix. Performance of the system is very dependent on how much physical memory (RAM) you have. If you don't have enough RAM to run your workload, performance degrades rapidly.
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 space
RMCmem includes a simple command to summarize this:
# /opt/RMCmem/bin/prtmem
Total memory:             989 Megabytes
Kernel Memory:             60 Megabytes
Application:              110 Megabytes
Executable & libs:         42 Megabytes
File Cache:               757 Megabytes
Free, file cache:          11 Megabytes
Free, free:                 6 Megabytes
Total physical memory
The total physical memory can be seen using prtconf. Memory is allocated in units called pages, and you can use the 'pagesize' command to see the size in bytes per page:
# /usr/sbin/prtconf | grep Memory
Memory size: 1024 Megabytes
# /usr/bin/pagesize
Kernel memory
Kernel memory is allocated to hold the initial kernel code at boot time, then grows dynamically as new device drivers and kernel modules are used. Kernel tables also grow dynamically, unlike some older versions of Unix. As you add hardware and processes to a system, the kernel will grow. In particular, to keep track of all the memory in a system, the kernel allocates a page table structure.
If you have several gigabytes of RAM this table gets quite large. The dynamic kernel memory allocator grabs memory in large "slabs," then allocates smaller blocks more efficiently. This means that the kernel tends to grab a bit more memory than it's really using. If there is a severe memory shortage, the kernel unloads unused kernel modules and devices and frees unused slabs. The simplest summary of kernel memory usage comes from sar. To show the kernel memory allocation (KMA) activities use (see man sar for more details).
# sar -k 1
SunOS diamond 5.7 Generic_106541-12 sun4u    04/28/01
sml_mem   alloc  fail   lg_mem    alloc  fail  ovsz_alloc  fail
6873088 6044236     0 44818432 43761720     0    11231232     0
Application process memory
Application processes consist of an address space divided into segments, where each segment maps either to a file, anonymous memory (the swap space), System V shared memory, or a memory mapped device. The mapped files include the code and 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 command in the RMCmem package.
# /opt/RMCmem/bin/pmem 361
361:    ora_dbw0_DIA3
  Kbytes Resident Shared Private Permissions       Mapped File
   82992   82992   82992       - read/write/exec   [shmid=0x2]
      16      16       8       8 read/exec
      16      16       8       8 read/exec
       8       8       8       - read/write/exec
........      ..      ..       . ...............   ...........
     112      80      72       8 read/exec
       8       8       8       - read/write/exec
      16      16       8       8 read/exec
       8       8       8       - read/write/exec
--------  ------  ------  ------  ------
  124232   93040   92728     312
Now we can see that the process address space size is 124232 kilobytes; 93040 kilobytes of that are currently resident in main memory, wherein 92728 kilobytes are shared with other processes while 312 kilobytes are private. When this command started only 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 the shared memory for each mapped file, we'll know how much application memory is in use. This summary is shown by prtmem as we saw in the beginning, and the detail is listed by the memps command in RMCmem.
# /opt/RMCmem/bin/memps
PID     Size Resident   Shared  Private  Process
...  .......   ......   ......     ....  .............
359  118904k   93608k   92800k     808k  ora_pmon_DIA3
367  118184k   93152k   92704k     448k  ora_smon_DIA3
369  117928k   93120k   92704k     416k  ora_reco_DIA3
371  118040k   93136k   92720k     416k  ora_snp0_DIA3
365  119040k   93120k   92712k     408k  ora_ckpt_DIA3
377  118344k   93080k   92720k     360k  ora_d001_DIA3
363  119088k   93056k   92720k     336k  ora_lgwr_DIA3
375  118344k   93048k   92720k     328k  ora_d000_DIA3
361  124232k   93040k   92728k     312k  ora_dbw0_DIA3
373  121608k   93032k   92728k     304k  ora_s000_DIA3
Filesystem cache memory
This is the part of memory that is most confusing, as it is invisible. You can only tell it's there if you access the same file twice and it is quicker the second time.
The RMCmem package adds kernel instrumentation that counts up all the pages for each cached file. The memps -m command lists the files that are cached in order of the amount of memory they're consuming.
One problem is that within the kernel, the file is only known by its inode number and filesystem mount point. The directory pathname 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 the vnode open code) and making an inode-to-name lookup cache in the kernel. This cache size is limited (to 8192 entries by default), and the file may have been opened before the kernel module was loaded, so it can't always find the name.
# memps -m
   Size   InUse E/F Filename
 21064k  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
More infos about the RMCmem package can be found here as PDF

Using Sun Solaris Manuals directly from CD-ROM

Solaris  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)


Installation Instructions will cover the installation of the DLT tape peripheral hardware and configuration of the system to communicate 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 DLT drive can be attached only to a SCSI single-ended interface. The same is true for SCSI differential attachment. Solaris includes a driver to efficiently 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 scsi-bus using good 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 disk drives will never let you achieve any good throughput. You bought the DLT because of performance didn't 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!
boot -rv
Boot the system and log in as root. When booting you should see a message similar to these: "st1: ".
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=" tape unit>"," reference name>",""
  • tape-config-list is a variable defined by a series of tape configuration parameters listed below:
  • 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 and product ID for
      as described in the following table:
    DLT Tape Product
    DLT7000 QUANTUM DLT7000 (Total string character count, including spaces, must equal 15).
  • is a name you select that the system will use to identify the DLT device. This reference does not change the DLT product ID. When the system boots, the reference name will be displayed in the list of peripheral devices recognized by the system.
  • -data> is a variable containing a series of additional DLT device configuration information. You select a name in place of the string. You will continue editing the st.conf file by defining the name you selected for . The definition depends on the DLT tape peripheral you are installing. For a DLT7000 series unit add the following line:

contains 10 parameters and are described following:
The first parameter, is the version number and should not change.
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 */
The third parameter is the block size. Since the DLT tape drive uses variable block size, this value should be zero.
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
F                       0x0008
                      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.
The fifth parameter, 4, defines the number of densities. The maximum definable number of densities is 4.
The sixth, seventh, eighth and ninth parameter are used for system selection of tape densities. Use these values for a DLT 7000 Tape Drive.
The tenth parameter defines which density the system will use as the default density. The sixth, seventh, eighth and ninth parameters in the 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:

shutdown-i0-g0 boot -rv
Reboot the System
The -r switch in the boot command enables a kernel compile and includes the creation of device special files used for communication with the DLT device. The -v switch enables verbose mode display of system bootup. With verbose mode, the system should indicate that the DLT tape peripheral is attached by displaying the string you selected.
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= 0

The target drive designations assigned by Solaris may take on values higher than already established in the /dev/rmt/ path. This is not a problem but during a boot -rv, Solaris does not remove tape device files for drives that are no longer attached to the system. This can increase the effort in locating the device file for the configured drive, however, this can be minimized by first deleting the tape device files:
rm /dev/rmt/*
then either boot the system with a:
boot -rv
or issue the following at the command line prompt:
drvconfig -i st; tapes
If the DLTtape is the only drive on the system, it's target assignment should be zero. The Solaris man pages have more information on drvconfig and tapes.

  Reconfigure Devices on Solaris

If you remove or add a device on Solaris then the devices files must be recreated, either with boot -rv or devfsadm. For example to renumber the logical tape drive devices do the following: Tape drives were numbered beginning with /dev/rmt/3 instead of /dev/rmt/0. The physical devices pointed to by the logical /dev/rmt/[012] devices no longer existed, and we wanted to renumber 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 -v

devfsadm(1M) maintains the /dev and /devices namespaces. It replaces the previous suite of devfs administration tools including drvconfig(1M), disks(1M), tapes(1M), ports(1M), audlinks(1M), and devlinks(1M).
The following options are supported:
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 values for device_class: disk, tape, port, audio, and pseudo. This option may be specified more than once to specify multiple device classes.

  OpenBoot Diagnostics

The Solaris operating system gets the jumpstart for its booting from a hardware-level interface called the OpenBoot PROM or OBP for short. OpenBoot at its heart has an interactive command interpreter with a varied set of functions. OBP is a firmware which is stored in the socketed startup PROM of the computer and consists of two parts, the PROM and the NVRAM.
As stated earlier while the PROM acts as the interface for access to diagnostics and drivers, the NVRAM consists of some editable user defined parameters. Non Volatile information like the system identification information, device aliases etc are stored in the NVRAM.The OpenBoot PROM is programmable and can be programmed based on Forth, which is an interactive
programming 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 boot
Useful commands at OK prompt.
Dignostics : boot  General
this command shows the following  systems hardware informatiion : Model, architecture, processor,keyboard, openboot version, Serial no. ethernet  address & host id.

test floppy - test floppy disk drive
test net - test network loopbacks
test scsi - test scsi interface
test-all    test for all devices with selftest method

Show ticks of real-time clock

Monitor network broadcast packets

Monitor broadcast packets on all net interfaces

Show attached SCSI devices
Show attached SCSI devices for all host adapters- internal & external.
boot - boot kernel from default device.
Factory default is to boot
from DISK if present, otherwise from NET.

boot net - boot kernel from network
boot cdrom - boot kernel from CD-ROM
boot disk1:h - boot from disk1 partition h
boot tape - boot default file from tape
boot disk myunix -as - boot myunix from disk with flags "-as"
ok cd /pci@1f,4000/scsi@3 
ok .properties
ok ls
f00809d8 tape
f007ecdc disk
ok .speed
CPU Speed : 200.00MHz
UPA Speed : 100.00MHz 
PCI Bus A : 66Mhz
PCI Bus B : 33Mhz
printenv Display all variables and current values.
Set variable to the given   value.
Reset the value of variable to the factory default.
Reset variable values to the factory defaults.
Key Sequences
These commands are disabled if the PROM security is on. Also, if your system has full security enabled, you cannot apply any of the suggested commands 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 to start 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

OK setenv diag-switch? true OK setenv auto-boot? false OK reset-all
OK test-all or obdiag
Configure Graphics Console (e.g. Sun XVR-100 Graphics Accelerator) instead of serial TTYA
OK show-displays
Select the graphics accelerator, e.g. b
OK nvalias mydev
OK setenv output-device mydev OK setenv use-nvramrc? true OK reset-all

Why doesn't my .forward file work?

If you are having problems where you have created a $HOME/.forward file in your home directory to forward e-mails from one account to another and it just won't forward them?
Set correct Permissions
First make sure the file isn't 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   # OK
Lastly, make sure your home directory isn't group or world writable.
drwxrwxr-x 14 zahn dba 4096 Jan 9 12:20 zahn   # wrong drwxr-xr-x 14 zahn dba 4096 Jan 9 12:20 zahn   # OK

Simple Shell Script to backup your Files

A backup strategy is more complex than creating a redundant copy of disk storage and considering the strategy a success. A successful backup strategy must detail how the backup media are rotated, how the media are archived, how the system will be recovered, and what the backup software will do to create the backup. Although all parts of the backup strategy are equally important, this tip will focus on the backup script and will detail a flexible backup script that uses built-in Solaris software tools which create a reliable local backup of a Solaris machine.
The 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 system.
  • The backup script will run autonomously. The only human intervention will be to swap media and review output.
  • The filesystems or directories to backup can be specified in the script. Using automounter you can even 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 media.
  • The backup script will be able to run on any Solaris 2.6 or greater machine without modification.
Tools used
We use the well known utilities TAR, GZIP and DD, because they are available on any Unix system. They are very well tested and simple to use. In case of an emergency it is important to have a simple way to restore, independent of complex tools and incompatible software releases.
Magnetic Tape Control
The utility MT sends commands to a tape drive. Many of these commands are familiar, but some are not. The script will use these mt commands.
  • 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
Click here for the Shell Script