Thursday 16 August 2007

Unix topics

split from chunk and download
split -b 100m nidb.dmp nidb_

find text in all files
find . -type f -exec grep "\"select" {} \; -print | grep jsp

mount share via nfs (this is softshare, has a time)
mount -o soft lnx-nidb1.jhq:/u02/oracle /mnt/nidb1

Unix mail from command line
echo "myTest body" | mail -s "myTest subject" nislam@sch.com

Format disk
with ext3
mkfs.ext3 /dev/sda1

Restarting netowrk
service network restart

Plug the USB then check dmesg to see device location e.g. sdg: sdg1 sdg2
mount /dev/sdg1 /mnt/usbdisk

List open file (fuser / lsof)
============
fuser system01.dbf
/usr/sbin/lsof system01.dbf

Find old files and delete (30 days)
find /u01/oracle/NIDB/arch -maxdepth 1 -mtime +30 -name "*.arc" -exec rm -rf {} \;

Change OS time
cd /etc
ln -sf /usr/share/zoneinfo/Etc/GMT localtime

But most things will need a restart before it takes affect, can we find some time to action a reboot?

Check Netowrk speed
ethtool eth0


Database Disk Layout
Good discussion on db disk and Raid
http://asktom.oracle.com/pls/asktom/f?p=100:11:0::::P11_QUESTION_ID:359617936136


New Database (2*74) (12*300)
Here are disk layout I plan to implement on new db box.

Internal (2x 74GB) Raid 1
/boot 150MB
/ 10GB
/home 5GB
Swap 16384MB
/u01 15GB Ora Software, admin
/u02 25GB Redo 1, Undo

External (6x 300GB) Raid 1+0
Swap 8192MB
/u11 150GB App Datafile
/u12 150GB App Datafile
/u13 100GB App Datafile
/dbbackup1 500GB (Or rest) RMAN backup and Exp dmp

External (4x 300GB) Raid 1+0
Swap 8192MB
/u21 100GB Index
/u22 100GB Index
/u23 100GB Ineex
/dbbackup2 300GB (Or rest) RMAN backup

External (2x 300GB) Raid 1
/u31 10GB Redo 2, Temp
/u32 40GB Redo 3, Arc
/u33 250GB (Or rest) Other

If I had more disks I would have moved Redo 1 from OS / Software disk.

No comments: