Home » RDBMS Server » Server Administration » Is ASM a solution to a large recovery area (11.2.0.1 SE RH 7.1)
| Is ASM a solution to a large recovery area [message #645023] |
Mon, 23 November 2015 21:20  |
 |
juniordbanewbie
Messages: 250
Registered: April 2014
|
Senior Member |
|
|
Dear all,
currently I have a system in which all the files are scattered across different filesystem/partitions,
from the OS, the mount point is as follow
[root@chicago-db ~]# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/rhel_chicago--db-root 22G 3.3G 19G 16% /
devtmpfs 16G 0 16G 0% /dev
tmpfs 16G 0 16G 0% /dev/shm
tmpfs 16G 25M 16G 1% /run
tmpfs 16G 0 16G 0% /sys/fs/cgroup
/dev/sdd 355G 301G 36G 90% /u01
/dev/sdg 355G 161G 176G 48% /u05
/dev/sde 355G 303G 34G 91% /u02
/dev/sdb 355G 303G 34G 90% /u04
/dev/sdf 355G 309G 28G 92% /u03
/dev/sdh 837G 483G 312G 61% /import2
/dev/sdc 1008G 89M 957G 1% /import
/dev/sda1 497M 120M 378M 24% /boot
from the db, the breakdown of database file size is as follow:
SYS@camden>SELECT substr(file_name,1,4), sum(bytes)/1024/1024/1024 size_in_g FROM dba_data_files GROUP BY ROLLUP(substr(file_name,1,4)) ORDER BY sum(bytes)/1024/1024/1024;
SUBS SIZE_IN_G
---- ----------
/u05 40.2929688
/u01 249.622993
/u04 278.001892
/u02 278.322205
/u03 283.998962
1130.23902
conclusion recovery area for database should be about at least twice the size of current database file size which means recovery area should be about 2260 GB
if it is not possible to have a filesystem of 2260 GB, I guess the only solution is to have ASM
from http://docs.oracle.com/cd/E11882_01/server.112/e18951/asmcon.htm#OSTMG94059
Quote:
If you omit the failure group specification, then Oracle ASM automatically places each disk into its own failure group, except for disk groups containing disks on Oracle Exadata cells.
Normal redundancy disk groups require at least two failure groups. High redundancy disk groups require at least three failure groups. Disk groups with external redundancy do not use failure groups.
My interpretation of the above is as follow:
normal redundancy disk group can have 3 or even more disks because it state at least. so for external redundancy we can have as many disk as possible right?
So all I need to do is as follow
if each partition can only be 355G, about 7 disk 2260/355 are required, then I need to partition and configure the Disk Devices to Use the Oracle Automatic Storage Management Library Driver according to http://docs.oracle.com/cd/E11882_01/install.112/e47689/oraclerestart.htm#CHDFEDHB
install Oracle Grid Infrastructure for a Standalone Server 11.2.0.4 (Binaries for ASM) from p13390677_112040_Linux-x86-64_3of7.zip
Create the Diskgroup FRA
Migrate the existing flash recovery to FRA
Right?
Is there any other issues we have?
thanks and many thanks
[Updated on: Mon, 23 November 2015 22:11] Report message to a moderator |
|
|
|
| Re: Is ASM a solution to a large recovery area [message #645024 is a reply to message #645023] |
Mon, 23 November 2015 23:09  |
 |
Mahesh Rajendran
Messages: 10708
Registered: March 2002
Location: oracleDocoVille
|
Senior Member
Account Moderator |
|
|
>>if it is not possible to have a filesystem of 2260 GB
Why would it not be possible?
Behind all the glory, ASM is just a volume manager. You can treat it as you like it.
Unless you are planning to have FRA in local disks / Oracle Database Appliances (as opposed to SAN/Arrays which are already raided,
having dedicated controlllers etc and have decent Unix support to maintain them),
You can just use external redundancy.
If you are planning to use ASM for FRA, migrate the database as well into ASM.
Else, it would be an overkill.
[Updated on: Mon, 23 November 2015 23:20] Report message to a moderator |
|
|
|
|
|
|
|
| Re: Is ASM a solution to a large recovery area [message #645032 is a reply to message #645023] |
Tue, 24 November 2015 01:08 |
John Watson
Messages: 8976
Registered: January 2010
Location: Global Village
|
Senior Member |
|
|
One more point -
Quote:I need to partition and configure the Disk Devices to Use the Oracle Automatic Storage Management Library Driver
This is not a good idea. The ASMLib is history, replaced in 12.1.0.2 with the ASM Filter Driver. However, the FIlter Driver is (I think) not yet ready for production use. I would advise that you use ether at the moment, but rater use udev to configure the devie ownership and modes. udev is easy and reliable and your system administrators love it.
|
|
|
|
|
|
| Re: Is ASM a solution to a large recovery area [message #645034 is a reply to message #645033] |
Tue, 24 November 2015 01:51 |
John Watson
Messages: 8976
Registered: January 2010
Location: Global Village
|
Senior Member |
|
|
|
At the moment, the only way to install the Filter Driver is to install ASMLib first. Which is stupid. I am sure beyond all reasonable doubt that this will be fixed in a future release. Which is why I would use neither now, and stay with udev.
|
|
|
|
| Re: Is ASM a solution to a large recovery area [message #645041 is a reply to message #645026] |
Tue, 24 November 2015 05:53 |
|
Mahesh Rajendran
Messages: 10708
Registered: March 2002
Location: oracleDocoVille
|
Senior Member
Account Moderator |
|
|
>> why is it overkill?
Because ASM is just a volume manager, with certain special implementation tailored to be used by an Oracle database.
Why would you go through all the pain (setting up and maintaining the Grid) and use it just as filesystem for FRA?
FRA can happily reside outside, just like a conventional file system.
|
|
|
|
Goto Forum:
Current Time: Mon May 05 09:00:06 CDT 2025
|
] 
Blog
Search
Help
Members
Register
Login
Home
