Archive for March 18th, 2008

Zettabyte File System ZFS

Posted by carlosap in solaris on March 18th, 2008

Features:

1. 256 quadrillion zettabytes ( Terabytes, Petabytes, Exabytes, Zettabytes (1024 Exabytes))
2. RAID-0/1 & RAID-Z ( RAID-5 with enhancements ) ( 2 – required virtual devices )
3. Snapshots – read-only copies of file systems or volumes
4. Create volumes
5. Uses storage pools to manage storage – aggregates virtual devices
6. File systems attached to pools grow dinamically as storage is added
7. File systems may span multiple physical disk
8. ZFS is transactional (Similar to databases. Example: Writing 100 mb, but only 80mb is commited in a traditional file system, leaving 20 mb unwriting, and the data is corrupted, and in ZFS all is written or NONE.
9. Pools & file systems are auto-mounted. No need to maintain /etc/vfstab
10. Supports file system hierarchies: /pool1/{home(5GB),var(10GB), etc.}
11. Supports reservation of storage: /pool1(36GB)/{home(10GB),var(36–10GB)}
12. Provides a secure web-based management tool

*********** ZFS – CLI **************
# which zpool
/usr/sbin/zpool

zpool list – list known pools
zpool create pool_name(alphanumeric, _,-,:,.)

Pool Name Constraints
1. mirror
2. raidz

—— Pause: note to vmware —————————————————————-

  Add a second disk to the image in vmware

In order to add second hard disk with Fusion.

  • solaris must be halted.
  • VM must be shut down.
  • Click the + sign, add disk and enter a size.
  • devfsadm  (almost typed reboot — -r but that would be “old think” so that format sees the new device.)

format
Searching for disks…done

AVAILABLE DISK SELECTIONS:
       0. c1t0d0 <DEFAULT cyl 1302 alt 2 hd 255 sec 63>
          /pci@0,0/pci1000,30@10/sd@0,0
       1. c1t1d0 <DEFAULT cyl 2557 alt 2 hd 128 sec 32>
          /pci@0,0/pci1000,30@10/sd@1,0
———————————————————————————————————-

Continue….

zpool create pool_name device_name1, device_name2, device_name3, etc.

bash-3.00# devfsadm
bash-3.00# format
Searching for disks…done

AVAILABLE DISK SELECTIONS:
       0. c0d0 <DEFAULT cyl 2607 alt 2 hd 255 sec 63>
          /pci@0,0/pci-ide@7,1/ide@0/cmdk@0,0
       1. c0d1 <DEFAULT cyl 1303 alt 2 hd 255 sec 63>
          /pci@0,0/pci-ide@7,1/ide@0/cmdk@1,0
Specify disk (enter its number): ^Z

[1]+  Stopped                 format
bash-3.00# zpool create pool1 c0d1
bash-3.00# echo $?
0

mount
/pool1 on pool1 read/write/setuid/devices/exec/xattr/atime/dev=2d50002 on Tue Mar 18 09:24:51 2008–

bash-3.00# zpool list
NAME                    SIZE    USED   AVAIL    CAP  HEALTH     ALTROOT
pool1                  9.94G     88K   9.94G     0%  ONLINE     -

ZFS Pool Statuses
1. ONLINE
2. DEGRADED (part of the mirror is broken or entire disk is fail, but still available)
3. FAULTED
4. OFFLINE
5. UNAVAILABLE

(Pools or datasets)

—————————-
zfs list – returns ZFS dataset info

bash-3.00# zfs list
NAME    USED  AVAIL  REFER  MOUNTPOINT
pool1  85.5K  9.78G    25K  /pool1

zfs mount – returns pools and mount points
bash-3.00# zfs mount
pool1                           /pool1

zpool status – returns virtual devices that constitute pools
bash-3.00#zpool status
  pool: pool1
 state: ONLINE
 scrub: none requested
config:

        NAME        STATE     READ WRITE CKSUM
        pool1       ONLINE       0     0     0
          c0d1      ONLINE       0     0     0

errors: No known data errors

bash-3.00#zpool status -v pool1

Note: ZFS requieres a minimum of 128MB virtual device to create a pool

(Recommendation render all the disks over zfs minus /  )
var home should be separate disks for performance

——————-DESTROY ————————————————
zpool destroy pool1 – Destroys pool and associated file system

bash-3.00# zpool destroy pool1
bash-3.00# echo $?
0

—————-CREATE File systems within pool1 ———————
zfs create pool1/home – creates file system named ‘home’ in pool1

bash-3.00# zfs create pool1/home
bash-3.00# echo $?
0

bash-3.00# zfs list
NAME         USED  AVAIL  REFER  MOUNTPOINT
pool1        114K  9.78G  25.5K  /pool1
pool1/home  24.5K  9.78G  24.5K  /pool1/home

Note: Default action of  ‘zfs create pool1/home’ assigns all storage available to ‘pool1’, to ‘pool1/home’

##### Set quota on existing file system ######
zfs set quota=5G pool1/home

bash-3.00# zfs set quota=5G pool1/home
bash-3.00# zfs list
NAME         USED  AVAIL  REFER  MOUNTPOINT
pool1        114K  9.78G  25.5K  /pool1
pool1/home  24.5K  5.00G  24.5K  /pool1/home

##### Create user-based file system beneath pool1/home #####

zfs create pool1/home/zivo

bash-3.00# zfs get -r quota pool1
NAME                    PROPERTY  VALUE               SOURCE
pool1                     quota       none                 default
pool1/home             quota       5G                    local
pool1/home/pedrito   quota       none                default
pool1/home/zivo       quota       none                default

zfs get -r compression pool1
NAME                PROPERTY     VALUE               SOURCE
pool1               compression  off                 default
pool1/home          compression  off                 default
pool1/home/pedrito  compression  off                 default
pool1/home/zivo     compression  off                 default

############ Rename file systems ######################

bash-3.00# zfs rename pool1/home/zivo pool1/home/zivob
bash-3.00# echo $?
0

############ Extending dynamically, pool storage ######################
AVAILABLE DISK SELECTIONS:
       0. c0d0 <DEFAULT cyl 2607 alt 2 hd 255 sec 63>
          /pci@0,0/pci-ide@7,1/ide@0/cmdk@0,0
       1. c0d1 <VMware V-0000000000000000-0001-10.00GB>
          /pci@0,0/pci-ide@7,1/ide@0/cmdk@1,0
       2. c1d1 <DEFAULT cyl 4093 alt 2 hd 128 sec 32>
          /pci@0,0/pci-ide@7,1/ide@1/cmdk@1,0

bash-3.00# zpool add pool1 c1d1
bash-3.00# zfs list
NAME                     USED  AVAIL  REFER  MOUNTPOINT
pool1                    210K  17.6G  25.5K  /pool1
pool1/home              76.5K  5.00G  27.5K  /pool1/home
pool1/home/pedrito      24.5K  5.00G  24.5K  /pool1/home/pedrito
pool1/home/zivokickass  24.5K  5.00G  24.5K  /pool1/home/zivokickass

bash-3.00# zpool status
  pool: pool1
 state: ONLINE
 scrub: none requested
config:

        NAME        STATE     READ WRITE CKSUM
        pool1       ONLINE       0     0     0
          c0d1      ONLINE       0     0     0
          c1d1      ONLINE       0     0     0

errors: No known data errors

 

No Comments

no hostkeys solaris 10 sshd

Posted by carlosap in solaris on March 18th, 2008 

cat /var/svc/log/network-ssh\:default.log


Could not load host key: /etc/ssh/ssh_host_rsa_key
Could not load host key: /etc/ssh/ssh_host_dsa_key



cd /etc/sshssh-keygen -d -f ssh_host_dsa_key -N ""ssh-keygen -b 1024 -f ssh_host_rsa_key -t rsa -N ""


 


 


 

No Comments

Solaris 10 HANDBOOK

Posted by carlosap in solaris on March 18th, 2008

Solaris 10

  • admintools is not implemented
  • pkgchk -P instead of grep /var/sadm/install/contents
  • RBAC: Role-Based access control (kind of built in sudo)

Solaris management commands

smc Starts the Solaris management console
smcront Manage cron jobs
smdiskless Manage diskless client support
smexec Manage entries in the exec_attr database
smgroup Manages group entries
smlog Manages and views WBEN log files
smmultiuser Manages bulk operations on multiples accounts
smosservice Add OS services and diskless client support
smprofile Manages profiles in the prof_attr and exec_attr db
smrole Manages roles and users in role accounts
smserialport Manages serial ports
smuser Manages user entries

how to Reboot / Halt / Poweroff

On Solaris using the shutdown command to reboot, halt or power off your server is the best option, because this command will call the rc script.

Reboot

shutdown -i6 -g60 -y

-i : init level
-g : grace time (60 secondes in this case)
-y : Pre-answer yes

Halt

shutdown -i0 -g60 -y

Power Off

shutdown -i5 -g60 -y

How to send a break on Solaris 10 x86

You will have to boot the system with “kadb”.

From the command line

# eeprom boot-file=kadb
# reboot

OR

From the GRUB boot prompt choose edit and edit the second line

root (hd0,0,a)
kernel /platform/i86pc/multiboot kadb
module /platform/i86pc/boot_archive

Drop to kadb, generate a core file

From a directly attached keyboard or serial port connection, type: 

 F1-A - press the "F1" and "A" keys, simultaneously.

You can test it to make sure it’s work and then just enter: 

:c

to return to the normal mode.

The control-alt-d key sequence also works.

Once at the kadb prompt, type: 

$<systemdump

This will generate a core file, which can be retrieved from: 

 /var/crash/'uname -n' (or wherever the local system stores core files)

How to determine 32-bit or 64-bit?

How to determine whether a system has 32-bit or 64-bit solaris capablilities enabled 

# isainfo -v
32-bit sparc applications

How to display the native instruction sets executable?

Isalist prints the names of the native instruction sets executable on this platform on the standard output 

# isalist
sparcv8 sparcv8-fsmuld sparcv7 sparc

How to displays information about processors?

# psrinfo -v
Status of processor 0 as of: 01/24/07 13:03:35
  Processor has been on-line since 01/19/07 14:13:09.
  The sparc processor operates at 60 MHz,
        and has a sparc floating point processor.

How to find out if the the third processors is up?

if [ "`psrinfo -s 3 2> /dev/null`" -eq 1 ]
     then
          echo "processor 3 is up"
     fi

How to display solaris relase information?

# cat /etc/release
                       Solaris 8 10/00 s28s_u2wos_11b SPARC
           Copyright 2000 Sun Microsystems, Inc.  All Rights Reserved.
                            Assembled 31 August 2000

How to display a system’s installed memory?

# ptrconf
System Configuration:  Sun Microsystems  sun4m
Memory size: 128 Megabytes
System Peripherals (Software Nodes):

SUNW,SPARCstation-20
...

Howto show machine and software revision information

# showrev
Hostname: clust01
Hostid: 7235260f
Release: 5.8
Kernel architecture: sun4m
Application architecture: sparc
Hardware provider: Sun_Microsystems
Domain:
Kernel version: SunOS 5.8 Generic 108528-03 August 2000

Howto show patched installed list

# showrev -p | more
Patch: 109965-01 Obsoletes:  Requires:  Incompatibles:  Packages: SUNWpamsc
Patch: 109618-01 Obsoletes:  Requires:  Incompatibles:  Packages: SUNWeuxwe, SUNWeuezt, SUNWeudlg, SUNWeudda
Patch: 108725-02 Obsoletes: 109579-01 Requires:  Incompatibles:  Packages: SUNWcsu, SUNWcsr, SUNWhea
Patch: 108825-01 Obsoletes:  Requires:  Incompatibles:  Packages: SUNWcsu
Patch: 108875-07 Obsoletes:  Requires:  Incompatibles:  Packages: SUNWcsu, SUNWcsr, SUNWcsl, SUNWarc, SUNWcstl, SUNWhea
Patch: 108899-01 Obsoletes:  Requires:  Incompatibles:  Packages: SUNWcsu
Patch: 108982-05 Obsoletes: 109440-01 Requires:  Incompatibles:  Packages: SUNWcsu, SUNWhea
Patch: 108985-02 Obsoletes:  Requires:  Incompatibles:  Packages: SUNWcsu
Patch: 109043-02 Obsoletes:  Requires: 109041-01 Incompatibles:  Packages: SUNWcsu
Patch: 109045-02 Obsoletes:  Requires: 109041-01 Incompatibles:  Packages: SUNWcsu

How to display locales installed on a system (solaris 10)

# localeadm

How to determine if a local is installed on a system (solaris 10)

# localeadm -q

Display host id number

# hostid
723560f

How to display the user ownership of local UFS

# quot -a
/dev/rdsk/c0t1d0s0 (/):
834291  root
312013  bin
  812   uucp
  146   adm
   91   lp
   22   nobody
    4   daemon
    3   listen
/dev/rdsk/c1t1d0s0 (/opt):
538104  root
66742   bin
 2168   #3792
/dev/rdsk/c0t1d0s7 (/export/home):
    9   root

Howto configuring Telnet / FTP to login as root?

Now before getting into the details of how to configure Solaris for root logins, keep in mind that this is VERY BAD security. Make sure that you NEVER configure your production servers for this type of login.

Configure Telnet for root logins

Simply edit the file /etc/default/login and comment out the following line as follows: 

# If CONSOLE is set, root can only login on that device.
# Comment this line out to allow remote login by root.
#
# CONSOLE=/dev/console

Configure FTP for root logins

First remove the ‘root’ line from /etc/ftpusers.
Also, don’t forget to edit the file /etc/ftpaccess and comment out the ‘deny-uid’ and ‘deny-gid’ lines. If the file doesn’t exist, there is no need to create it.

NOTE: If you are using Solaris 9 or Solaris 10, the ftp* files are located in /etc/ftpd

How to setup timezone?

/etc/TIMEZONE

Howto setup nis client?

  • set the domain name 
domainname nis.domaine.com
  • edit /etc/hosts
#
# Internet host table
#
127.0.0.1       localhost

# NIS
192.168.1.100 nis1.domain.com
192.168.1.101 nis2.domain.com
  • ypinit
ypinit -c

pkgadd common mistake

In this example I’m trying to install the package “SUNWbash” but it could be any package, you may have been through those different cases: 

# pkgadd SUNWbash                                        
pkgadd: ERROR: no packages were found in </var/spool/pkg>
 
# pkgadd ./SUNWbash                                        
pkgadd: ERROR: no packages were found in </var/spool/pkg>
 
# pkgadd /mnt/Solaris_10/Product/SUNWbash       
pkgadd: ERROR: no packages were found in </var/spool/pkg>
 
# pkgadd -d ./SUNWbash
pkgadd: ERROR: no packages were found in </mnt/Solaris_10/Product/SUNWbash>

In the three first cases “pkgadd” looks in the default folder “/var/spool/pkg” in order to find packages and install them but none is found. In the fourth case the package himself is defined as a folder containing packages, that’s why it fails.
The right solution is:

# pkgadd -d /mnt/Solaris_10/Product/ SUNWbash
or
# pkgadd -d $PWD SUNWbash

Thus pkgadd will consider “/mnt/Solaris_10/Product/” as default folder and “SUNWbash” as package to install.

Where is the packages folder

/var/sadm/pkg/

Package installed

/var/sadm/install/content

Which package owns an installed file

pkgchk -l -p /usr/bin/ls

If you cannot access your sun box, ssh/ftp/console whatever

Then try and issue a break from the console. 

telnet> sent brk
Type  'go' to resume
{0} ok
{0} ok boot -s
Resetting ...

OBP Quick Memo

SEND A BREAK

From a telnet session: 

ctrl+]
send brk

From a ssh session: 

~break

Quick memo

help (show categories)

help sys (show sys help)

printenv (varible name and status)

setenv (to change)

setenv auto-boot? true

banner (show the mac address)

prob-scsi-all (find all scsi devices, show number off harddrives)

boot -s ( single user mode)
boot cdrom -s (boot single user mode from the cdrom)
boot net (boot network)

boot net:dhcp - install (Boot install using jumpstart with DHCP)

boot net - install (Boot install using jumpstart with Bootp)

boot cdrom - install (Boot install using the cdrom)

devalias (same as /etc/path_to_inst)

devalias toto /sbus/SUNW,fdtwo

show-devs (show all dev)

show-nets (show net interface)

set-default

From the running os: 

eeprom

No Comments

mount cdrom solaris core install : p

Posted by carlosap in solaris on March 18th, 2008

 Unless you’ve changed the system’s configuration, mounting
CD/DVD media is supposed to happen automatically, with the
help of the “vold” daemon.

The CD/DVD will be mounted somewhere in /cdrom/xxxx with a
symbolic link /cdrom/cdrom0 pointing to the root directory of the
media.

If you’ve disabled “vold”, of if you didn’t install it, manual mounting
works like this:

1. iostat -En
This command can be used to print a list of SCSI/ATAPI/USB/
Firewire disk/cdrom/dvdrom devices. Note the disk id in front
of the entry, for example “c1t1d0″:

% iostat -En
c1t1d0           Soft Errors: 1 Hard Errors: 0 Transport Errors: 0
Vendor: HL-DT-ST Product: DVDRAM GMA-4020B Revision: A105 Serial No:
Size: 0.00GB <0 bytes>
Media Error: 0 Device Not Ready: 0 No Device: 0 Recoverable: 0
Illegal Request: 1 Predictive Failure Analysis: 0

2. Manually mount the device, using the device id from the above
iostat command, with slice “s2″ appended (on Solaris x86:
append “p0″).

# mount -F hsfs /dev/dsk/c1t1d0s2 /mnt

——SOLARIS X86–—-mount -F hsfs /dev/dsk/c1t0d0p0 /mnt

No Comments