Custom Logrotate in Solaris 10

Here I explain how to configure logadm to rotage any system wide files according to given criteria.
1. Add the corresponding entries in /etc/logadm.conf in below format.
root@server1 # tail -3 /etc/logadm.conf
/var/adm/wtmpx -A 1m -o adm -g adm -m 664 -p 1d -t '$file.old.%Y%m%d_%H%M' -z 1
/var/adm/wtmpx -A 1m -g adm -m 664 -o adm -p 1w -t '$file.old.%Y%m%d_%H%M' -z 5
/var/adm/utmpx -A 1m -g adm -m 664 -o adm -p 1w -t '$file.old.%Y%m%d_%H%M' -z 5
/var/adm/loginlog -A 1m -g sys -m 700 -o root -p 1w -t '$file.old.%Y%m%d_%H%M' -z 5
Explanation for each switch:
-A ->Delete any versions that have not been modified for the amount of time specified by age. Specify age as a number followed by an h (hours), d (days), w(weeks), m (months), or y (years).
-o -> the owner of the newly creating empty file
-g-> the group of newly creating file
-m ->mode of the new empty file (chmod xxx)
-p -> Rotate a log file after the specified time period (period as d, w,  m, y)
-t -> Specify the template to use when renaming log files (Here, wtmpx.old.20101225_0757) (see man logadm for more info)
-z  ->How many copy of rotaged files needs to retain on the system.
-P ->Used by logadm to record the last time the log was rotated in /etc/logadm.conf (no need to set this manually)
2. Once above entries are done, execute logadm -v command to run a logrotation now. Now logadm reads the /etc/logadm.conf file, and for every entry found in that file checks the corresponding log file to see if it should be rotated.
root@server1 # logadm  -v
# loading /etc/logadm.conf
# processing logname: /var/log/syslog
#     using default rotate rules: -s1b -p1w
#     using default template: $file.$n
# processing logname: /var/adm/messages
#     using default rotate rules: -s1b -p1w
#     using default template: $file.$n
# processing logname: /var/cron/log
#     using default expire rule: -C10
# processing logname: /var/lp/logs/lpsched
#     using default rotate rules: -s1b -p1w
# processing logname: /var/fm/fmd/errlog
#     using default expire rule: -C10
#     using default template: $file.$n
# processing logname: /var/fm/fmd/fltlog
#     using default template: $file.$n
# processing logname: smf_logs
#     using default template: $file.$n
# processing logname: /var/adm/pacct
#     using default template: $file.$n
# processing logname: /var/log/pool/poold
#     using default expire rule: -C10
#     using default template: $file.$n
# processing logname: /var/svc/log/system-webconsole:console.log
#     using default rotate rules: -s1b -p1w
#     using default expire rule: -C10
#     using default template: $file.$n
# processing logname: /var/opt/SUNWsasm/log/sasm.log
#     using default template: $file.$n
# processing logname: /var/adm/wtmpx
mkdir -p /var/adm # verify directory exists
mv -f /var/adm/wtmpx /var/adm/wtmpx.old.20101225_1250 # rotate log file
touch /var/adm/wtmpx
chown adm:adm /var/adm/wtmpx
chmod 664 /var/adm/wtmpx
#     recording rotation date Sat Dec 25 12:50:51 2010 for /var/adm/wtmpx
# processing logname: /var/adm/utmpx
mkdir -p /var/adm # verify directory exists
mv -f /var/adm/utmpx /var/adm/utmpx.old.20101225_1250 # rotate log file
touch /var/adm/utmpx
chown adm:adm /var/adm/utmpx
chmod 664 /var/adm/utmpx
#     recording rotation date Sat Dec 25 12:50:51 2010 for /var/adm/utmpx
# processing logname: /var/adm/loginlog
mkdir -p /var/adm # verify directory exists
mv -f /var/adm/loginlog /var/adm/loginlog.old.20101225_1250 # rotate log file
touch /var/adm/loginlog
chown root:sys /var/adm/loginlog
chmod 700 /var/adm/loginlog
#     recording rotation date Sat Dec 25 12:50:51 2010 for /var/adm/loginlog
# writing changes to /etc/logadm.conf
As you can see the last line of above command, once the logadm command successfully run, it do some changes to with -P switch in /etc/logadm.conf file regarding the last update of logrotation.
root@server1 # tail -3 /etc/logadm.conf
/var/adm/wtmpx -A 1m -P 'Sat Dec 25 12:50:51 2010' -g adm -m 664 -o adm -p 1w -t '$file.old.%Y%m%d_%H%M' -z 5
/var/adm/utmpx -A 1m -P 'Sat Dec 25 12:50:51 2010' -g adm -m 664 -o adm -p 1w -t '$file.old.%Y%m%d_%H%M' -z 5
/var/adm/loginlog -A 1m -P 'Sat Dec 25 12:50:51 2010' -g sys -m 700 -o root -p 1w -t '$file.old.%Y%m%d_%H%M' -z 5
List of new files created in /var/adm
root@server1 # ls -ltr /var/adm/*.old*
-rwx------   1 root     sys            0 Dec 25 11:00 /var/adm/loginlog.old.20101225_1250
-rw-r--r--   1 root     bin         3720 Dec 25 15:49 /var/adm/utmpx.old.20101225_1250
-rw-rw-r--   1 adm      adm      8595060 Dec 25 15:51 /var/adm/wtmpx.old.20101225_1250

Colon Separated WWPN in Soalris

The script prints the WWPN of Solaris 10 HBA cards WWPN in a familiar colon separated format that Storage Admin can use easily in SAN devices.

root@solaris-10 # fcinfo hba-port|grep 'HBA Port WWN'|awk {'print $4'}|perl -pe 's/(..)/\1:/g'|cut -c 1-23
21:00:00:1b:32:95:56:2c
21:01:00:1b:32:b5:56:2c
21:00:00:1b:32:95:66:2c
21:01:00:1b:32:b5:66:2c

Downloading Soalris 10 Recommended patch cluster

Click here for Link

Solaris zfs raw device

Here I am creating a zfs raw device for using as raw partition for Sybase database.
The list of available zpool.

bash-3.00# zpool list
NAME     SIZE   USED  AVAIL    CAP  HEALTH  ALTROOT
dbpool   19.9G  1.32G  18.6G     6%  ONLINE  -
Creating a test volume of 10MB. (Notice the -V option with zfs command, which create a tree for corresponding device under /dev/zvol. This is option is required only when you want to appear this device under /dev/zvol)

bash-3.00# zfs create -V 10M dbpool/test
Creating UFS file system for the created device
bash-3.00# newfs /dev/zvol/rdsk/dbpool/test
newfs: construct a new file system /dev/zvol/rdsk/dbpool/test: (y/n)? y
Warning: 4130 sector(s) in last cylinder unallocated
/dev/zvol/rdsk/dbpool/test:      20446 sectors in 4 cylinders of 48 tracks, 128 sectors
        10.0MB in 1 cyl groups (14 c/g, 42.00MB/g, 20160 i/g)
super-block backups (for fsck -F ufs -o b=#) at:
32,
Now the newly created zfs volume is available in below list

bash-3.00# zfs list
NAME              USED  AVAIL  REFER  MOUNTPOINT
dbpool            1.33G  18.2G    21K  legacy
dbpool/apps         24K  10.0G    24K  /app
dbpool/lotus      1.32G  18.2G  1.32G  /opt/lotus
dbpool/notesdata    21K  18.2G    21K  /notesdata
dbpool/noteslogs    21K  18.2G    21K  /noteslogs
dbpool/test         10M  18.2G    24K  -
Its ready to mount it now.
bash-3.00# mount /dev/zvol/dsk/dbpool/test /mnt
bash-3.00# df -h |grep dbpool
dbpool/lotus             20G   1.3G    18G     7%    /opt/lotus
dbpool/notesdata         20G    21K    18G     1%    /notesdata
dbpool/noteslogs         20G    21K    18G     1%    /noteslogs
dbpool/apps              10G    24K    10G     1%    /app
/dev/zvol/dsk/dbpool/test   7.5M   1.0M   5.7M    16%    /mnt
Below is the path for zfs raw device and block device.

Block Device=/dev/zvol/dsk/dbpool/test
Raw Device=/dev/zvol/rdsk/dbpool/test


Ref:

Starting and Accessing LDOM

Listing all the available domains

bash-3.00# ldm ls
NAME             STATE      FLAGS   CONS    VCPU  MEMORY   UTIL  UPTIME
primary          active     -n-cv-  SP      4     1G       2.3%  14h 30m
orange1          inactive   ------          4     8G
orange2          inactive   ------          4     8G


Now both LDOMS are inactive. For activating the same, we have to bind it first.


bash-3.00# ldm bind-domain orange1


Now it can be  booted
bash-3.00# ldm start-domain orange1


Accessing the LDOM console using telnet to port 5000 ( for next domain it will be 5001 and so on)


bash-3.00# telnet 0 5001
Trying 0.0.0.0...
telnet: Unable to connect to remote host: Connection refused
bash-3.00# telnet 0 5000
Trying 0.0.0.0...
Connected to 0.
Escape character is '^]'.


Connecting to console "orange1" in group "orange1" ....
Press ~? for control options ..
Hostname: orange1
Jun 19 12:07:07 in.mpathd[198]: missed sending 1 probes cur_time 52464930 snxt_time 52466262 snxt_basetime 52465390
/dev/rdsk/c0d0s4 is clean
/dev/rdsk/c0d0s5 is clean
/dev/md/rdsk/d80 is clean
Reading ZFS config: done.


orange1 console login:

Project for Solaris 10 Oracle installation

This is the simple and default project for solaris 10 oracle installation:

add below lines in /etc/project file

group.dba:100:oracle setting::dba:process.max-sem-nsems=(privileged,2048,deny);project.max-sem-ids=(priv,100,deny);project.max-shm-ids=(priv,100,deny);project.max-shm-memory=(priv,11811160064,deny)

Here:

this is effective for all users in dba system user group

100 is the ID for this project

oracle setting is the description for this prject

you can verify this valule using project - l command