©2015 - 2022 Chad’s Technoworks. Disclaimer and Terms of Use

Chad’s TechnoWorks My Journal On Technology

Information Technology

Oracle 12c ASM Installation and Configuration - page 2

Prev< 1 2 3 >Next

Disk Group for spfile:

  click diskgroup button to create a diskgroup called RECO_DG with the following raw disks as normal redudancy:

  asmreco01_c1t1d0s6

  asmreco02_c1t2d0s6


Listener: ASM_LISTENER  <<--- The listener we previously created using the netca


ASM parameter:

defaults are shown on screen shot asmca_3

replace default with recommended values:

memory_target = 1536m

processes = 335

After applying the changes, proceed to create ASM.


After the initial diskgroup and instance creation, you'll be presented with Disk Group Management Window where there is only one disk group in the list - RECO_DG. You then add more disk groups based on the available raw disks. In my case, I'll add DATA_DG where I intend to store all database files related to tables and indexes. I also specified the Failure Groups A and B in pairs. Ideally, you should have at least two hardware disk controlers for redundancy purposes. In order for your mirroring to be effective, you have to group the raw disks with the same controllers so that group "A" will have disks belonging to controller 1 and Group "B" will have disks of controller 2.

This setup will help isolate controller failures besides the potential disk failures.

In my case I only have one controller "c1" since this is just a lab.


Failure Group A

  asmdata01_c1t3d0s6

  asmdata02_c1t4d0s6

Failure Group B

  asmdata03_c1t5d0s6

  asmdata04_c1t6d0s6

Exit asmca and we are done. Your ASM is now ready to be used by your databases.


Some important things to keep in mind,


               Processes = 25 + (10 + max number of concurrent database file creations, and file extend operations possible) * n

                    where n = the number of databases connecting to ASM



HOW TO CHECK THE STATUS OF DISK GROUPS


SQL> set linesize 200

SQL> column group format a20

SQL> column path format a35

SQL> column failgroup format a20

SQL> column disk format a20

SQL> select g.name "Group", d.failgroup "Failgroup", d.name "Disk", d.mode_status "Status", d.path "Path"

  2  from v$asm_disk d, v$asm_diskgroup g

  3  where d.group_number = g.group_number;


Group                Failgroup            Disk                 Status  Path

-------------------- -------------------- -------------------- ------- -----------------------------------

RECO_DG              RECO_DG_0000         RECO_DG_0000         ONLINE  /dev/rdsk/asmreco01_c1t1d0s6

RECO_DG              RECO_DG_0001         RECO_DG_0001         ONLINE  /dev/rdsk/asmreco02_c1t2d0s6

DATA_DG              A                    DATA_DG_0000         ONLINE  /dev/rdsk/asmdata01_c1t3d0s6

DATA_DG              A                    DATA_DG_0001         ONLINE  /dev/rdsk/asmdata02_c1t4d0s6

DATA_DG              B                    DATA_DG_0002         ONLINE  /dev/rdsk/asmdata03_c1t5d0s6

DATA_DG              B                    DATA_DG_0003         ONLINE  /dev/rdsk/asmdata04_c1t6d0s6


6 rows selected.


SQL>  


HOW TO ADD DISKS TO A DISK GROUP


First, Let's determine the path of raw disks of your environment.

The default raw disk path of the attached storage devices of your server host are listed below according to the type of operating environment.

     OS Platform           Raw Disk Path

     --------------------  ---------------

     Solaris (32/64 bit)   /dev/rdsk/*

     Windows NT/XP         \\.\orcldisk*

     Linux (32/64 bit)     /dev/raw/*

     Linux (ASMLIB)        ORCL:disk*

     HPUX                  /dev/rdsk/*

     HPUX(Tru 64)          /dev/rdisk/*

     AIX                   /dev/rhdisk/*

 

The ASM init.ora parameter has the info on where to find these raw disks.


SQL> show parameter asm_disk


NAME                                 TYPE        VALUE

------------------------------------ ----------- ------------------------------

asm_diskgroups                       string      DATA_DG

asm_diskstring                       string      /dev/rdsk/asm*

SQL>


The above query result tells us that the ASM should be looking for raw disks having an alias starting with the word 'asm' as stated by the asm_diskstring.


In my case, I had introduced additional raw devices to this path having the same naming convention.

You can determine if these new raw disks (labeled with CANDIDATE) are seen by the ASM using the following query.


SQL> column path format a35

SQL> select path, header_status, mode_status from v$asm_disk;


PATH                                HEADER_STATU MODE_ST

----------------------------------- ------------ -------

/dev/rdsk/asmdata05_c1t7d0s6        CANDIDATE    ONLINE

/dev/rdsk/asmdata06_c1t8d0s6        CANDIDATE    ONLINE

/dev/rdsk/asmreco01_c1t1d0s6        MEMBER       ONLINE

/dev/rdsk/asmreco02_c1t2d0s6        MEMBER       ONLINE

/dev/rdsk/asmdata01_c1t3d0s6        MEMBER       ONLINE

/dev/rdsk/asmdata02_c1t4d0s6        MEMBER       ONLINE

/dev/rdsk/asmdata03_c1t5d0s6        MEMBER       ONLINE

/dev/rdsk/asmdata04_c1t6d0s6        MEMBER       ONLINE


8 rows selected.


SQL>