Server Administration :: Disk Group With Two Disk For Normal Redundancy
Jan 16, 2013
I Configured an ASM instance and a disk group with two disk for normal redundancy.
> Here .. each disk is 2gb
The disk group has two disks...
SQL> select group_number, name, type, total_mb, free_mb
2 from v$asm_diskgroup;
GROUP_NUMBER NAME TYPE TOTAL_MB FREE_MB
------------ ------------------------------ ------ ---------- ----------
1 DATA NORMAL 4000 3898
as the group has two way mirroring (Normal redundancy) How much data (2 GB or 4 GB) can i keep in the disk group? My conception is I can keep 2 GB data in the disk group... (as the disk group keeps every extent in another disk as mirror)
One of my collegues created an ASM disk group with normal redundancy using 2 disks. But the disks are not of same size, one disk is 100GB and another one is 10GB. Now the usable space out of diskgroup is showing 55GB. When I checked the diskgroup properties it is showing 2 fail groups one is DATA1_ 0000 with 10GB and other is DATA_0001 with 100GB. My question is why it is showing 55GB as usable space? My assumption is as it is having 2 fail groups with different disks of different sizes. For the failgroup 2 even it is of 100GB size, in order to maintain the redundancy with other small FG(10G) it will consider only 10GB out of 100GB. So, the 2nd FG size also should be 10GB. So that the usable space should show as 10GB as opposed to 55GB (not (100+10)/2).
My cluster have two node.Data diskgroup uses normal redundancy.Priamry disks are located in one storage.Fail group disks are located in another storage.When we close the failgroup disks' storage the 2nd instance restart.After automatically restarted everything goes well.I want to know why does the 2nd instance restart?Is it normal?
We have a Production Oracle 10g R2 RAC on HP-UX v2 IA64 servers.We have Two Disk Groups one for Archive (ARC_DISK - 100 GB) and other for Database(DATA_DISK - 1 TB]. We wanted to add more space to the DATA_DISK disk group.Unix admin configured 200 GB from SAN and changed the ownership of the Disk to oracle and permissions to 775 on 1st Node.I opended DBCA from 1st Node and was able to see the disk in 'Show Candidate'.
I added this disk to the DATA_DISK disk group and clicked OK but got ORA- error with some message like some operations could not be performed. I exited DBCA.We realized that we had forgotten to change the ownerhip and permission from the 2nd Node.Unix admin changed the ownership of the Disk to oracle and permissions to 775 on the 2nd Node.
I opened DBCA again from 1st Node and selected the DATA_DISK disk group but could not find the Disk in 'Show Candiate' open. I clicked on 'Show All' and this disk was shown with Header_Status - MEMBER but not allocated to DATA_DISKGROUP. When I clicked the 'Show Member' option, this disk is not shown for DATA_DISK disk group. I exited DBCA at this point.As this is a critcal Production database I didnt proceed any further and exited DBCA.
Now I need to add this Disk to the DATA_DISK disk group but not sure which option to select. I got one reply from another forum to run DBCA select the DATA_DISK Disk Group and then click 'Show All' and select this Disk (which already has MEMBER as Header Status) and select Force Option and click OK to continue.
I was setting up disks groups and I accidentally created one group (DATA) with "NORMAL" redundancy but wanted it to be "EXTERNAL". I tried using asmca to remove disks from the group, drop the group, change the redundancy..... All of this failed because there was an spfile on the disk group.
I finally got it to work with using this procedure:
sqlplus '/ as sysasm' SQL*Plus: Release 11.2.0.3.0 Production on Thu Apr 5 08:58:19 2012 Copyright (c) 1982, 2011, Oracle. All rights reserved. SQL> drop diskgroup DATA; drop diskgroup DATA * ERROR at line 1: ORA-15039: diskgroup not dropped ORA-15053: diskgroup "DATA" contains existing files
[code]....
In summary, I am not sure why changing the redundancy would be so difficult if there is data on the disk group.
I have 2 servers both having windows server 2008 64 bit as operating system installed on both I need to install oracle clusterware 11g r1 on both servers with clustering on external storage. I have configured the network(private,public and virtual) for both servers and have started the installation.
In the installation of oracle I add both servers but then I reach to a point where they ask me for voting disk or ocr disk in the cluster configuration storage but no disk is present how can i create ocr disk or voting disk on windows server 2008? And the external storage should I buy a special type of storage that supports clustering to continue my work?
I checked and found we have disk that is assigned with 0 disk GROUP_NUMBER. What does that mean ? how to check if disk T1_ASM05 is been part of any disk group or not.?
SQL> select GROUP_NUMBER,NAME from v$asm_diskgroup;
GROUP_NUMBER NAME ------------ ------------------------------ 1 DATA 2 FRA SQL> SQL> select GROUP_NUMBER,name,PATH from v$asm_disk;
What is best practice to change small disk D:? I am beginner with Oracle. 10g on W2008. 5 datafiles (all indexes,second data file, 2 undotabs)*.dbf (34;30;1;34;12 GB) is on D:. Part of tablespaces (1 data, 1 undo)has files on c:.
I. 1.Shutdown 2008 server. 2.Copy D: image with GHOST to USB, network. 3.Connect new D, create RAID. 4.Restore image to D. 5.Start 2008 server.
II. 1.Stop application. 2.CONNECT AS SYSDBA 3.SHUTDOWN NORMAL or (IMMEDIATE)? 4.Copy files *.dbf at OS level from d: to ... USB disk, network. 5.Shutdown 2008 server. 6.Change disks, create RAID in BIOS. 7.Start W2008. Is Oracle at this moment in SHUTDOWN mode? 8.Copy back *.dbf to new D: (with directory structure). 9.STARTUP Oracle.
What should be our approach when we see the disk response time is bad for a particular tablespace in database.I heard a good disk response time should be on an average 10ms.
According to my understanding , if Disk1 Fails Disk4 facilitates normal operations. When there is space crunch it operates in reduced redundancy . Am i right ?
2.I have got 4 Disks in one group (i.e from Disk1 To Disk4 ) i have not defined any failure group and as per my understanding all disks will be added to its own failure group without mirroring and striping.
I have a small problem resizing an ASM disk group . I've had a lun with size 50G and added on it an asm disk called LUN1D1. Then I have created a DISK GROUP called data1 (external) witch contain LUN1D1. DATA1 was then made asm volume. So I have an asm volume with size 50G and I need it to be 100G.
Now I have increased the size of the lun with 50G (100G in total). I have rescanned the LUN and os see the new size. The problem is I can not re size DISK GROUP data1 or disk LUN1D1.
essentially create data fragmentation within the datafile resulting in the db having lots more space to write into but not actually freeing space, even if you shrink the file it doesnt free space or do a reorg?
We have as an example a DB with 2 billion rows of data in 1 table, no partioning just one large table.
We have worked out that we can probably delete 1 billion rows or even better only keep a rolling 3 month window of data.
What would be the suggestion on deleting this data and reclaiming the disk space to actually see additional disk space made available at the os level.
deleting the data and reclaiming the space.
Through reading it looks like it might be something like, delete, creating new table space partitions from this data. This in theory would create new a tablespace in newly created data files which would result in the data being reorganised and taking up less physical space and when completed you point to the newly created partitions and drop the old tables.
My understanding of DB_FILE_MULTIBLOCK_READ_COUNT parameter is that it affects only Full Table Scans and Fast Full Index Scans - all other disk retrieval is single block.If so, then maybe I'm reading this trace incorrectly:
select /*+ first_rows */ pk from test_join_tgt where pk >= 0 and rownum > 1
We are using Oracle 10g and have 10 tablespaces defined for our Database which have 108 tables. Size of 108 tables is around 251 MB as seen during importing the dump. While creating these 10 tablespaces I used below parameters for allocation of space
SIZE 1M REUSE AUTOEXTEND ON NEXT 1M MAXSIZE 1M;
which set the initial space for 10 tablespaces to around 1032Kb each. Now my Question is after importing the dump , how the disk space for 10 tablespaces increases to 398 MB in total ?
Is there any relation of Tablespace disk space and Actual Data present in the tables ?
i am trying to identify the pro's/con's of using multiple ASM Diskgroup. I understand oracle recommends/best practice is to have 2 DG (one data and one flash) and you can place multiple copies of control files/online redo logs(and thats the way i want to go). But would that same be true if i use different set of DISK. For example we have multiple RAID 10 devices and multiple of SSD devices for us that we can use for this ASM instance. And i was thinking to create 2 more Disk group (call it DG_SYS1 and DG_SYS2) and use that to put my online redo logs, control file and temp and system table space there. i understand in a standalone system(where regular file system is being used), they(online redo/ control file) are usually on there own drives, but with ASM when i am already using external RAID 10 config + ASM striping i assume the IO would faster or am i better of using the SSD that i can have for my redo/control? What would be the pro's/cons of it (besides managing multiple DG)..
I have 11.1.0.7 instance running and got more 300G of storage available for the database. I would like to know what are the steps to add these disks/storage to the existing DATA asm disk group?
- one ASM instance - X DB instances - each DB instance uses 2 or more dedicated diskgroups from the ASM instance - there is one diskgroup named FREEDISK that contains spare disks
On each DB instance you can see:
- the list and global parameters of all diskgroups using v$asm_diskgroup view - the list and parameters of all disks the instance is using with v$asm_disk view
So my question is: how (if this is possible) to know the list of (spare) disks in FREE DISK disk group?
, I have a few doubts in Normal,High and External redundancy levels in ASM concept. External - No failure group. No mirroring. Normal - Two Way Mirroring. One failure group. High - Three Way Mirroring. Two failure group. 1.
above mentioned 3 types in correct.2. My main question is what is the minimum number of disks needed in Disk group creation in Normal and External redundancy??
I managed to upload images to a database server, resize them, copy to the application server and everything worked just fine - the Apex page successfully displayed images. Since last week, things have broken. This is how: there's a directory object which points to application server's directory:
SQL> select * from all_directories;
OWNER DIRECTORY_NAME DIRECTORY_PATH ------- ------------------------------ ----------------------------------- SYS SLIKE_4005_UPLOAD d:gisslike_4005_upload --> on a database server SYS SLIKE_4005 \my-iasd$homegisslike_4005 --> on an application server
SQL>
I can use a directory located on a database server:
D:GISSlike_4005_upload>dir photo_resize.* Volume in drive D is RAID Volume Serial Number is 88F2-69D2 Directory of D:GISSlike_4005_upload [code]....
How come it doesn't work? I was absent last week, database server was restarted for some reason (there were Windows' updates which required restarting). After that, all applications (lucky us, just two of them, but in multiple procedures/functions) return FALSE for UTL_FILE.FGETATTR.
We recreated directory objects, but that didn't work (UNC or not, no difference). I Googled quite a lot, read Metalink notes - nothing I did solved the problem.
what these OS updates were about; maybe they are not to be blamed at all. Both servers (database & application) run MS Windows Server 2003 Standard Edition Service Pack 2. In the meantime, a colleague developed a workaround (it uses UTL_HTTP) which works, but it is MUCH slower than the previous UTL_FILE.FGETATTR option.
Why don't we keep these images on the database server (instead of the application server)?I was told that Apache is incapable of accessing mapped network directories so we used what we could.
Lets say I have an oracle setup on C: and I decided to delete the oracle in c: and setup a whole new oracle at E: but I want all the data migrate to the new oracle installation in E: How can I perform that?
I have a question about the ADD VOLUME command, I can't understand the difference between ADD DISK and ADD VOLUME.What are the difference between them?When should I use each one?How can I control the stripping and mirroring (NORMAL and HIGH) adding VOLUMES in a DISKGROUP?Can I add a volume to a fail group?
I'm trying to create a ASM disk using oracleasm.I created list the disk (using oracleasm), but the view V$ASM_DISK and ASMCMD (lsdsk command) can't see it..Look:
We have a new implementation which will be using ASM with RAID. The data area needs to be 3TB, and the recovery area, to be used for archive logs and RMAN backups, needs to be 1TB.
The configuration i'm thinking about now is:
+DATA diskgroup: 5*600G disks, using RAID 1+0 (this will include control files) +FRA diskgroup: 2*500G disks, using RAID 1 +LOG diskgroup: 2*1G disks, using RAID 1 (this is only for redo logs)
So here are my questions: 1. Am I right in supposing that we would get the best performance on the +FRA and +LOG diskgroups by not using RAID 1+0? 2. It has already been agreed to use RAID 1+0 for the +DATA diskgroup, but I can't see the added benefit of this. ASM will already stripe the data, so surely RAID 0 will just stripe it again (i.e double striping the data). Would it not be the better just to mirror the data at a hardware level?
I just made the swich to have my RMAN backup directly to tape and here are the parameters I have set.
CODECONFIGURE DEFAULT DEVICE TYPE TO 'SBT_TAPE'; CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FOR DEVICE TYPE SBT_TAPE TO '%F'; CONFIGURE DEVICE TYPE 'SBT_TAPE' PARALLELISM 4 BACKUP TYPE TO BACKUPSET; CONFIGURE DATAFILE BACKUP COPIES FOR DEVICE TYPE SBT_TAPE TO 1; # default
[code]...
When I do an active clone I need to set my parameters back to the following in order to makeit work.
CODECONFIGURE DATAFILE BACKUP COPIES FOR DEVICE TYPE DISK TO 1; # default CONFIGURE ARCHIVELOG BACKUP COPIES FOR DEVICE TYPE DISK TO 1; # default CONFIGURE MAXSETSIZE TO UNLIMITED; # default CONFIGURE ENCRYPTION FOR DATABASE OFF; # default
[code]....
Instead of un-doing my RMAN TAPE settings, is there something I can do in my active clone script to tell it read/write from/to disk instead of tape. I tried allocatiing disk channels but was unable to get my script to work. The only way it worked was to back out all the tape settings and put everything back to disk.
CODErman catalog=rman/rman@prod target=sys/sys@prod << EOT connect auxiliary sys/sys@dev duplicate target database to dev from active database nofilenamecheck pfile=/u01/app/oracle/product/11g/dbs/initdev.ora; exit
