How to add disk to the ASM
Usefull Link:
http://dba24.pl/2016/09/09/simple-procedure-of-adding-disks-to-asm-on-linux-single-instance/
PURPOSE: All documents are provided on this Blog just for educational purposes only. Please make sure that you run it in your test environment before to move on to production environment.
adding disks to ASM on Linux
Step1:
Add diskgroup
Login to the host.
ssh kardb.th.com
Check the existing disks.
ll /dev/mapper/*
ls /dev/mapper/kardb-lun_* | sort -n
or:
fdisk -l
# find wwid
lsscsi -i -s
Add new device in the multipath config file:
vi /etc/multipath.conf
multipath {
wwid 360050768018e02276000000000000c72
alias kardb-lun_56
}
Executing following commands.
rescan-scsi-bus.sh
or
rescan-scsi-bus.sh -a -w -m
service multipathd reload
ls /dev/mapper/kardb-lun_* | sort -n
or:
ls -al ls /dev/sd*
check the grants:
ll /dev/mapper/kardb-lun_53
or
ll /dev/sd*
If it is root:disk then set manuellay to oracle:dba
chown oracle:dba /dev/mapper/kardb-lun_53
If there is link check as well as follow:
ls -al /dev/mapper/kardb-lun_53
lrwxrwxrwx 1 root root 8 Nov 6 14:16 /dev/mapper/kardb-lun_53 -> ../dm-95
ls -al /dev/dm-95
brw-rw---- 1 oracle dba 253, 95 Nov 6 14:42 /dev/dm-95
Permissions ok. How do I do that it gets the owner:group automatically? By configuring udev here:
more /etc/udev/rules.d/99-oracle-asmdevices.rules
udevadm control --reload-rules && udevadm trigger
ENV{DM_NAME}=="kardb*",OWNER:="oracle",GROUP:="dba",MODE:="660"
Or another example:
KERNEL=="dm-*",ENV{DM_NAME}=="asm*p1", OWNER="grid", GROUP="asmadmin", MODE="0660"
su - oracle
asmcmd lsdsk -k -g --candidate
or:
sqlplus / as sysasm
select path from v$asm_disk where header_status='CANDIDATE';
sqlplus / as sysasm
alter diskgroup DATC add disk '/dev/mapper/kardb-lun_53';
Or normally no need add rebalance at the command. The above command manage automatically.
alter diskgroup DATC add disk '/dev/mapper/kardb-lun_53' rebalance power 3;
-- You can check running rebalance process.
select * from v$asm_operation;
SQL> exit
asmcmd lsdg
Completed.