- http://www.ugu.com - Unix Guru Universe
- http://www.MagniComp.com/sysinfo/ - SysInfo home page

Collecting system information:

Hostname:       % hostname Hostname aliases:       % grep `hostname` /etc/hosts | awk '{ print $3 }' Host network addresses:       % grep `hostname` /etc/hosts | awk '{ print $1 }' Host ID:       % hostid CPU type, application architecture, kernel architecture and version, operatin system name and version:       % uname -a Amount of main memory - can be found at boot time:       % dmesg Disk configuration (available space, etc.):       % df Other pieces:   A copy of all NFS mounted filesystems   NIS/YP configuration   A listing of all OS packages installed   A listing of all paches installed   Complete disk drive settings and configuration   All relevant software license keys and codes   Any custom soft links that were created   A copy of the setting and configuration for all defined printers

Backup some important files:

Kernel:   /kernel  or  /unix  or  /vmunix Passwd File:   /etc/passwd Group file:   /etc/group Host Table:   /etc/hosts Filesystem Table:   /etc/fstab  or etc/vfstab Sendmail Config Files:   /usr/lib/sendmail.cf, /usr/lib/sendmail.fc,  /usr/lib/sendmail.mc Inetd Configs:   /etc/inetd.conf TTY settings:   /etc/ttys Start Up Scripts:   files residing in /etc/init.d  or /etc/rc#.d directories

Executing script on the last day of the month (backups, filtering of log files, system utilization, etc.):

#!/bin/sh FILE = `runme` if test `TZ=EST -24 date +%d` = 1; then   $FILE fi

 

#!/usr/bin/perl use POSIX; @mydate = localtime(time); ++$mydate[3];  # add one day to the date $new_day_of _the_month = (localtime (POSIX::mktime (@mydate)))[3]; if ($new_day_of _the_month == 1) {    do_end_fo_the _month_stuff();  }

Using named pipes:

You can create named pipes using mknod or mkfifo commands, for example: #!/bin/sh mkfifo /tmp/testpipe echo "This is a test" > /tmp/testpipe & cat /tmp/testpipe rm /tmp/testpipe Here is a more elaborate example using 2 pipes to dump information from Oracle into a series of 10MB compressed files. $ mknod pipe1 p $ mknod pipe2 p $ nohup sqlplus user passwd @some_script_writing_into_pipe1 & $ nohup compress < pipe1 > pipe2 & $ nohup split -b 10245k < pipe2 &