Solaris Notes

• Default location of explorer conf file.

/etc/opt/SUNWexplor/default/explorer

• Where to e-mail files to have SUN look at them.

# Where explorer output should be mailed
EXP_EMAIL="explorer-database-americas@sun.com"

  Look for an installed package: pkginfo | openssl
  List all files in a package: pkgchk -l MAGENT | grep Pathname

  format - shows all disks, allows partitioning new disks
  iostat -En          # shows disk errorsexit

• Use format in script

  echo type 20 > format_opts # select proper HD type 20 is for EMC san
  echo label >> format_opts  # this will write the label to the disk
  format -f ./format_opts -d c4t5d101
  format -f ./format_opts -d c4t5d102
  format -f ./format_opts -d c4t5d103
  format -f ./format_opts -d c4t5d104
  format -f ./format_opts -d c4t5d105
  format -f ./format_opts -d c4t5d106
  format -f ./format_opts -d c4t5d107
  format -f ./format_opts -d c4t5d108
  

  /usr/platform/sun4u/sbin/prtdiag -v

  vxprint -thr        # shows volumes and plexes
  vxdisk list         # shows disk groups and disks
  vxdisk list <disk>  # shows disk info
  vxdctl
  vxdmpadm listctlr all   # list controllers
  format
  inq
  /usr/sbin/lpfc/lputil   # Emulex tool

  EMC Agent:  pkginfo -l MAGENT
  /usr/ecc/exec

  metastat -p     # shows a short configuration

        1. Find the bad disk
            iostat -En
            or
            dmesg
        2. Where is the metadb located?
            metadb -i
        3. If that disk is failing, move the metadb to only the good disk
            metadb -d /dev/dsk/c#t#d#s#         (bad disk)
            metadb -a -c 3 /dev/dsk/c#t#d#s#    (good disk)
        4. Physically replace the bad disk
        5. Copy the vtoc from the good mirror to the new disk
            prtvtoc /dev/rdsk/c0t0d0s2 | fmthard -s - /dev/rdsk/c0t0d0s2
            or
            format and
        6. Re-sync the mirror
            $ metastat | grep Invoke
                Invoke: metareplace d100 c0t1d0s0 <new device>
                Invoke: metareplace d101 c0t1d0s1 <new device>
                Invoke: metareplace d102 c0t1d0s7 <new device>
                Invoke: metareplace d103 c0t1d0s6 <new device>
                Invoke: metareplace d104 c0t1d0s5 <new device>

            Thus, the commands you need are:
            metareplace -e d100 c0t1d0s0
            metareplace -e d101 c0t1d0s1
            metareplace -e d102 c0t1d0s7
            metareplace -e d103 c0t1d0s6
            metareplace -e d104 c0t1d0s5

            Watch the re-sync
                $ metastat | grep %
                Resync in progress: 0 % done
                Resync in progress: 76 % done
                Resync in progress: 0 % done
                Resync in progress: 0 % done
                Resync in progress: 0 % done
        7. Restore the metadb
            $metadb -a -c3 /dev/dsk/c#t#d#s#
            $ metadb -i
                    flags           first blk       block count
                 a        u         16              1034            /dev/dsk/c0t0d0s3
                 a        u         1050            1034            /dev/dsk/c0t0d0s3
                 a        u         16              1034            /dev/dsk/c0t1d0s3
                 a        u         1050            1034            /dev/dsk/c0t1d0s3
                 a        u         2084            1034            /dev/dsk/c0t1d0s3
                 a        u         2084            1034            /dev/dsk/c0t0d0s3

  Userfull Commands:
      View print queue: lpstat -o printer.domain.net
      View printer status: lpstat -p printer.domain.net -l
      Stop a print queue: disable <printer-name>
      Start a pritn queue: enable <printer-name>
  Configuration info:
      /etc/printers.conf
      ypcat -k printer.conf.byname
      /etc/lp/interfaces/<printer-name>
  Log Files: /var/lp/logs
  Other Files: /var/spool/lp

• What process has is listening on specific port

# lsof -i TCP:7006
COMMAND   PID USER   FD   TYPE        DEVICE SIZE/OFF NODE NAME
inetd     827 root   18u  IPv4 0x300381944c0      0t0  TCP *:ctmagent (LISTEN) 
p_ctmat 26027 root    0u  IPv4 0x3001b3261d8      0t0  TCP dhextntc2.svr.bankone.net:ctmagent->ino6a003.svr.bankone.net:42775 (CLOSE_WAIT)

• How to add a Virtual IP

New VIP setup.
ifconfig ce1:28 plumb
ifconfig ce1:29 plumb
ifconfig ce1:28 155.180.250.170 netmask 255.255.254.0 broadcast 155.180.251.255 
ifconfig ce1:29 155.180.250.171 netmask 255.255.254.0 broadcast 155.180.251.255 
ifconfig ce1:28 up
ifconfig ce1:29 up

Modify Hosts file with the following:
155.180.250.170 pcsssvct.svr.bankone.net
155.180.250.171 pcsssvcdr.svr.bankone.net pcsssvcdr.img.bankone.net

#create hostname files with proper naming.
echo pcsssvct.svr.bankone.net > /etc/hostname.ce1:28
echo pcsssvcdr.svr.bankone.net > /etc/hostname.ce1:29

Using /net

cd /net/<hostname>/path

• list open file handles for a mount point

lsof | grep -i /{mount}/{point}

# lsof | grep -i /datahubt/u08
oracle    12359  oracle  261uW VREG   0,509542400                       /datahubt/u08 (/dev/vx/dsk/dhubt1/u08)
oracle    12359  oracle  262uW VREG   0,533725184                       /datahubt/u08 (/dev/vx/dsk/dhubt1/u08) 

• list active semaphores

# ipcs
IPC status from <running system> as of Tue Sep 13 15:33:57 EDT 2005
T         ID      KEY        MODE        OWNER    GROUP
Message Queues:
Shared Memory:
m          0   0x5e002a66 --rw-------     root     root
m          1   0x5643     --rw-rw-rw-     root     root
m          2   0x5654     --rw-rw-rw-     root     root
m        203   0xcace     --rw-rw-rw-     root    other
Semaphores:
s          0   0x1        --ra-ra-ra-     root     root
s          1   0x5653     --ra-ra-ra-     root     root
s          2   0x73657276 --ra-ra-ra-     root     root
s      65539   0x10057b2  --ra-ra-ra-     root     root
s     131076   0x4500097b --ra-ra-ra-     root    other
s          5   0x45000dca --ra-ra-ra-     root    other
s          6   0xcace     --ra-ra-ra-     root    other
s          7   0x450003a5 --ra-ra-ra-     root    other
s     262152   0x45000c22 --ra-ra-ra-     root    other
s          9   0x45000fa3 --ra-ra-ra-     root    other

• Command to remove a semaphore

NAME
     ipcrm - remove a message queue,  semaphore  set,  or  shared
     memory ID

SYNOPSIS
     ipcrm [ -m shmid ]  [ -q msqid ]  [ -s semid ]  [  -M shmkey
     ]  [ -Q msgkey ]  [ -S semkey ]

DESCRIPTION
     ipcrm removes one or more messages,  semaphores,  or  shared
     memory identifiers.

• Restart Tivoli Endpoint

LCFD Restart Procedure

1. ps -ef | grep engine | grep -v grep
img2s003n1 # ps -ef | grep engine | grep -v grep
    root  3060     1  0   Jun 29 ?       11:21 dm_ep_engine img2s003n1.svr.bankone.net solaris2 DM362317

2. run the tivoli stop script /etc/rc3.d/S**tiv
img2s003n1 # cd /etc/rc3.d/
img2s003n1 # ls
README               S69esmrc             S99vcs
S25mdlogd            S99Tivoli_lcf1       S99vvr-udp-settings
S50san_driverchk     S99netmask
img2s003n1 # ./S99Tivoli_lcf1 stop
The process id in /tivoli/lcf/dat/1/lcfd.pid is not an lcfd process id.
Unable to get process id of the Tivoli LCF daemon (pid).

3. kill engine process (verify that lcfd and tivoli are not running)
img2s003n1 # ps -ef | grep engine | grep -v grep
    root  3060     1  0   Jun 29 ?       11:21 dm_ep_engine img2s003n1.svr.bankone.net solaris2 DM362317
img2s003n1 # kill 3060
img2s003n1 # ps -ef | grep engine | grep -v grep

4. restart the script
img2s003n1 # ./S99Tivoli_lcf1 start
Aug 25 09:21:32 1 lcfd Command line argv[0]='/tivoli/lcf/bin/solaris2/mrt/lcfd'
Aug 25 09:21:32 1 lcfd Starting Unix daemon

5. verify that the engine comes up (give it a minute or so)

Below is the method to use if you hav a SPARC version of Solaris 8 and 9:

1. Get a copy of the Solaris 9 CD 1.
2. Do a 'boot cdrom -s'. This boots into single-user mode.
3. mount /dev/dsk/c0t0d0s0 /a
4. TERM=sun; export TERM
5. vi /a/etc/shadow/

Use vi to delete the password for root
Save shadow in vi: type ": wq!"

6. cd /
7. umount /a
8. cd /sbin
9. init 6 (this reboots you - log in as root, open up a terminal, type passwd, reset your passwd, and you're done!)