ASM diskgroup FREE_MB (v$ asm_diskgroup) incorrect information? (or how to recover the free space)

Similar Questions

• I asked for information on how to stop the automatic updates on my operating system to windows 7. I received information about windows 8.

  * Original title: stoppng updated Automatic windows 10

  I asked for information on how to stop the automatic updates on my operating system to windows 7.  I received information about windows 8.  Help, please.

  Change your windows update settings,

  Click on the Start button and type in the search box

  Windows updates

  Click on this list to the top

  

  

• I paid 9.99 for rent photoshop for a month.  How can I access it?  Adobe just bill you and give you any information on how to access the product. !

  I paid 9.99 for rent photoshop for a month.  How can I access it?  Adobe just bill you and give you any information on how to access the product. !

  Never mind. I thought about it.

• Difference between the free space on the disk and column FREE_MB in V$ ASM_DISKGROUP

  Hello
  
  in our RAC environment, we have setup an ACF. Mainly use us it for RMAN backups. By connecting to the ASM instance and run the query
  
  Select name, State, total_mb, free_mb from v$ asm_diskgroup;
  
  We get the following result:
  
  NAME STATE TOTAL_MB FREE_MB
  ------------------------------ ------------------- ----------------- ----------------
  10738 9912 MOUNTED DG00_CRS
  409592 360173 MOUNTED DG01_FR
  6553568 1628732 MOUNTED DG02_DATA
  DG03_BACKUP MOUNTED 5615920 4227
  
  Seems to me that the ASM instance is of the opinion that the amount of releases diskgroup DG03_BACKUP space is 4227 MB. BotR when checking on
  level of operating system, the result is different:
  
  1 K-blocks used available use % mounted on file system
  / dev/ASM/v03_backup-33
  5746196480 2430545120 3315651360/u01/app/oracle/orabackup 43%
  
  = > In fact, we have about 3.3 TB of free space available. Should this result will not appear when you run the above query? I currently have
  don't have an idea why the query returns 4227 MB...
  
  Any help will be appreciated...
  
  Rgds
  JH

  ACFS is a feature of 11.2, while the DSO has existed long before
  ASM will take into account the volumes such as space used when querying V$ ASM_DISKGROUP, charge the lsdg in asmcmd

  You can use the view V$ ASM_ACFSVOLUMES and V$ ASM_FILESYSTEM to ACFS
  or volinfo in asmcmd

• A bad synchronisation has replaced my profile with incorrect information. How can I get it back?

  I turned on sync on my computer at home without realizing that I implement a sync account for my work computer. Both computers have _very_ different bookmarks and login/password data. Before her, I realized, my home computer profile has been replaced with the profile of my work. My bad for the backup does not my profile, but here I am. I will lose physical access to that computer for four months on 10/23, so I'll try to get back to the former State. How can I do this? I am wary to try a reset, see how I screwed up my system.

  My original goal was to have two profiles on my work computer, one for home, for work, and the two synchronized according to the device name.

  A big thank you to all those who can help.

  Ben

  Normally the data would be merged between connected devices.

  You should be able to restore the bookmarks from a backup JSON (.json or tablet .jsonlz4) in the folder bookmarkbackups.

  • Bookmarks > show all bookmarks > import and backup > restore > choose file
  • http://KB.mozillazine.org/Backing_up_and_restoring_bookmarks_-_Firefox
  • https://support.Mozilla.org/KB/restore-bookmarks-from-backup-or-move-them

  If you have lost the password then try to see if you still the signons.sqlite (and key3db) in the Firefox profile folder to see if you can make Firefox reimport bookmarks from this file.

  32 of Firefox and later versions store the passwords in the file logins.json in the profile folder and no longer use the file signons.sqlite.
  If you have the file signons.sqlite which has been used in Firefox 31 and more, then you could try to recover the passwords stored in this file (you lose passwords saved in Firefox 32).
  You can force Firefox to re-import the passwords in the file signons.sqlite and regenerate the logins.json file.

  • reset the signon.importedFromSqlite pref on the topic: config page by default via the context menu
  • Delete the logins.json file in the closed Firefox with Firefox profile folder

  When you restart Firefox, then you should have the pref signon.importedFromSqlite with the value set to true and you should have the passwords into password manager, unless he could have errors.

  You can use this button to go to the Firefox profile folder currently in use:

  • Help > troubleshooting information > profile directory: see file (Linux: open the directory;) Mac: View in the Finder)

• How to recover the information about logon to outlook express

  I am using outlook express, but have forgotten my login data. How to recover?

  Contact your ISP. They know your logon information, as it is the same if you want to get your mail from their Web site.

• How to recover the Typekit account information

  A seller has purchased a Typekit fonts on our behalf, that we use on our site. I don't have account information to manage access to the police. We need to white list a new provider of this font. How can I get the account information associated with this policy which we bought?

  Hi Jeremy,.

  Please send us an email at [email protected] with the URL of your website and we can help you locate the Typekit account.

  Thank you!
  -liz

• How can I recover the disk space occupied by the uninstall information of Microsoft Update for all accepted updates.

  I would like to know how to do this for MS Windows XP Home and (especially) MS Windows Vista Home Premium (my sons PC is running out of space fast disk - I think because of updates).

  WinXP: See http://aumha.net/viewtopic.php?f=62&t=33827

  Vista: See http://aumha.net/viewtopic.php?f=77&t=33209 ~ Robear Dyer (PA Bear) ~ MS MVP (that is to say, mail, security, Windows & Update Services) since 2002 ~ WARNING: MS MVPs represent or work for Microsoft

• try to install KB955430 and get error 800f0825 can not find information on how to complete the upgrade

  KB 955430 is an important update, when I tried to install it, it failed.  The error message is 800f0825, but I can't find that he listed

  Hello

  see if that helps to fix:

  How to reset the Windows Update components?

  There is also an automatic 'fix - it' here

  http://support.Microsoft.com/kb/971058

  Also, try to put the KB numbers in the search on the link below and then manually download the

  http://www.Microsoft.com/downloads/en/default.aspx

   

  or please repost your question in the correct windows update forum

  http://answers.Microsoft.com/en-us/Windows/Forum/windows_vista-windows_update?page=1&tab=all

• How can I find a manual for a Dell Latitude E4300. I'm looking for more information on how to control the F keys on my computer.

  Where is a Dell latitude E4300 manual or manual in Windows Vista that can tell me how to turn on or off the F keys.  At the moment I get a few totally different symbols than usual.

  original title: activation or deactivation of the F keys

  Manuals are available from the

  http://support.Dell.com/support/eDOCS/systems/late4300/en/index.htm

• How to solve the problem of 'ASM disk group is inaccessible' slave VM rac2?

  People,
  
  Hello. I want to install Oracle 11 g RAC 2 using 2 virtual machines (VM main rac1 and rac2 VM slave) with Oracle Linux 5.6 according to the website http://appsdbaworkshop.blogspot.com/2011/10/11gr2-rac-on-linux-56-using-vmware.html
  
  Grid and data base RAC have been installed successfully for VMS rac1 and rac2.
  I use the slave rac2 VM test data grid and CARS because rac1 disk is full.
  The problem is that the OCR file is not available or currupted in the rac2 VM slave. To know his reasons, I check permission on disc ocr as below:
  
  [ora11g@rac2 bin] $ pwd
  / U02/11g_db/ora11g/racdb/bin
  
  [ora11g@rac2 bin] $ srvctl start asm
  
  RPRC-1070: could not verify if the ora.asm resource is registered
  Cannot communicate with crsd
  
  [ora11g@rac2 rac2] $ pwd
  / u01/app/Grid/log/Rac2
  
  [ora11g@rac2 rac2] $ cat alertrac2.log
  
  Last lines:
  [/ u01/app/grid/bin/cssdagent (5779)] CRS - 5822:Agent ' / u01/app/grid/bin/cssdagent_root "disconnected from the server. Details at (: CRSAGF00117 :) in u01/app/grid/log/rac2/agent/ohasd/oracssdagent_root/oracssdagent_root.log.)
  2012-03-20 13:44:55.914
  [ohasd (5149)] CRS-2112: the service ODE started on node rac2.
  2012-03-20 13:44:58.588
  [ohasd (5149)] CRS - 8017:location: / etc/oracle/lastgasp has 72 Advisory reboot log files, 0 were announced and 0 errors
  2012-03-20 13:45:07.555
  [ohasd (5149)] CRS - 2772:Server "rac2" was awarded to 'Free' the pool.
  2012-03-20 13:45:10.108
  [ohasd (5149)] CRS-2302: failed to get profile negotiators. Error CLSGPNP_NO_DAEMON (the demon GPNPD does not work).
  2012-03-20 13:45:10.251
  [gpnpd (5368)] CRS - 2328:GPNPD began the rac2 node.
  2012-03-20 13:45:13.037
  [cssd (5443)] CRS - 1713:CSSD daemon is launched in cluster mode
  2012-03-20 13:45:37.132
  [cssd (5443)] CRS - 1707:Lease acquisition for node rac2 number 2 completed
  2012-03-20 13:45:37.727
  [cssd (5443)] CRS - 1605:CSSD file to vote is online: ORCL:ASMDISK3; details in u01/app/grid/log/rac2/cssd/ocssd.log.
  2012-03-20 13:45:37.986
  [cssd (5443)] CRS - 1625:Node rac1, number 1, was closed manually
  2012-03-20 13:45:37.991
  [cssd (5443)] CRS - 1605:CSSD file to vote is online: ORCL:ASMDISK2; details in u01/app/grid/log/rac2/cssd/ocssd.log.
  2012-03-20 13:45:38.633
  [cssd (5443)] CRS - 1605:CSSD file to vote is online: ORCL:ASMDISK1; details in u01/app/grid/log/rac2/cssd/ocssd.log.
  2012-03-20 13:45:56.956
  [cssd (5443)] CRS - 1601:CSSD Reconfiguration is complete. Active nodes are rac2.
  2012-03-20 13:46:00.557
  [ctssd (5834)] CRS-2403: the time synchronization Service on rac2 host Cluster is in observer mode.
  2012-03-20 13:46:00.576
  [ctssd (5834)] CRS-2407: the new Cluster Time Synchronization Service reference node is host rac2.
  2012-03-20 13:46:01.000
  [ctssd (5834)] CRS-2401: time synchronization Service Cluster started on host rac2.
  2012-03-20 13:46:11.097
  [crsd (5856)] CRS-1013: the location of the OCR in ASM disk group is inaccessible. Details in u01/app/grid/log/rac2/crsd/crsd.log.
  2012-03-20 13:46:12.085
  [ohasd (5149)] CRS - 2765:Resource 'ora.crsd' failed on the server "rac2.
  2012-03-20 13:46:13.338
  [crsd (5925)] CRS-1013: the location of the OCR in ASM disk group is inaccessible. Details in u01/app/grid/log/rac2/crsd/crsd.log.
  2012-03-20 13:46:14.217
  [ohasd (5149)] CRS - 2765:Resource 'ora.crsd' failed on the server "rac2.
  2012-03-20 13:46:15.438
  [crsd (5940)] CRS-1013: the location of the OCR in ASM disk group is inaccessible. Details in u01/app/grid/log/rac2/crsd/crsd.log.
  2012-03-20 13:46:16.313
  [ohasd (5149)] CRS - 2765:Resource 'ora.crsd' failed on the server "rac2.
  2012-03-20 13:46:17.539
  [crsd (5955)] CRS-1013: the location of the OCR in ASM disk group is inaccessible. Details in u01/app/grid/log/rac2/crsd/crsd.log.
  2012-03-20 13:46:18.392
  [ohasd (5149)] CRS - 2765:Resource 'ora.crsd' failed on the server "rac2.
  2012-03-20 13:46:19.537
  [crsd (5988)] CRS-1013: the location of the OCR in ASM disk group is inaccessible. Details in u01/app/grid/log/rac2/crsd/crsd.log.
  2012-03-20 13:46:20.468
  [ohasd (5149)] CRS - 2765:Resource 'ora.crsd' failed on the server "rac2.
  2012-03-20 13:46:22.636
  [crsd (6003)] CRS-1013: the location of the OCR in ASM disk group is inaccessible. Details in u01/app/grid/log/rac2/crsd/crsd.log.
  2012-03-20 13:46:23.615
  [ohasd (5149)] CRS - 2765:Resource 'ora.crsd' failed on the server "rac2.
  2012-03-20 13:46:24.781
  [crsd (6022)] CRS-1013: the location of the OCR in ASM disk group is inaccessible. Details in u01/app/grid/log/rac2/crsd/crsd.log.
  2012-03-20 13:46:25.713
  [ohasd (5149)] CRS - 2765:Resource 'ora.crsd' failed on the server "rac2.
  2012-03-20 13:46:26.885
  [crsd (6037)] CRS-1013: the location of the OCR in ASM disk group is inaccessible. Details in u01/app/grid/log/rac2/crsd/crsd.log.
  2012-03-20 13:46:27.799
  [ohasd (5149)] CRS - 2765:Resource 'ora.crsd' failed on the server "rac2.
  2012-03-20 13:46:29.021
  [crsd (6052)] CRS-1013: the location of the OCR in ASM disk group is inaccessible. Details in u01/app/grid/log/rac2/crsd/crsd.log.
  2012-03-20 13:46:29.925
  [ohasd (5149)] CRS - 2765:Resource 'ora.crsd' failed on the server "rac2.
  2012-03-20 13:46:31.136
  [crsd (6067)] CRS-1013: the location of the OCR in ASM disk group is inaccessible. Details in u01/app/grid/log/rac2/crsd/crsd.log.
  2012-03-20 13:46:32.040
  [ohasd (5149)] CRS - 2765:Resource 'ora.crsd' failed on the server "rac2.
  2012-03-20 13:46:33.247
  [crsd (6082)] CRS-1013: the location of the OCR in ASM disk group is inaccessible. Details in u01/app/grid/log/rac2/crsd/crsd.log.
  2012-03-20 13:46:34.172
  [ohasd (5149)] CRS - 2765:Resource 'ora.crsd' failed on the server "rac2.
  2012-03-20 13:46:34.173
  [ohasd (5149)] CRS - 2771:Maximum attempts to restart the resource 'ora.crsd '; does not restart.
  [ora11g@rac2 rac2] $
  
  The output of alertrac2.log, we have seen disk group ASM is inaccessible. So my question is:
  
  How to solve the problem of "ASM disk group is inaccessible" in the rac2 VM slave?
  
  
  Thank you.

  user8860348 wrote:
  People,

  Hello. Thanks much for the reply.

  Yes. I put the entry "SQLNET. AUTHENTICATION_SERVICES = (NTS)"input file sqlnet.ora.

  I ran the commands below:

  [ora11g@rac2 bin] $ pwd
  / u01/app/Grid/bin

  [ora11g@rac2 bin] $ ps - ef | grep pmon

  16219 ora11g 8602-0 12:29 pts/2 00:00:00 grep pmon

  [ora11g@rac2 bin] $

  [ora11g@rac2 bin] $ export ORACLE_SID = + ASM2
  [ora11g@rac2 bin] $. / sqlplus / as sysdba

  SQL * more: Production version 11.2.0.1.0 Fri Mar 23 12:33:15 2012

  Copyright (c) 1982, 2009, Oracle. All rights reserved.

  Connect to an instance is idle.

  SQL > startup pfile = ' / u01/app/grid/dbs/pfile ".
  ORA-01031: insufficient privileges
  SQL >

  Sign in as SYSASM as

  . / sqlplus / as sysasm
  >

  As we see above shows the command $ ps - ef | grep pmon, my question is:

  First of all, are ' 16219 8602 ora11g 0 "process ID?

  Secondly, we need to kill one by one? For examples: kill-9 ora11g? Kill-9 16219?

  Laughing out loud. It's not the running process.

• ASM diskgroup space do not free themselves after RMAN archivelog delete

  Version Infra grid: 11.2.0.4

  DB version: 11.2.0.4

  OS: Oracle Linux 6.4

  Storage: Enterprise-class storage (EMC VMAX)

  We are in the Bank we have a database of CARS of Production 2 nodes very critical.

  Newspaper archives are stored in ARCH_DG diskgroup. Because of the unexpectedly heavy workload, our diskgroup ARCH_DG became full and we got the usual error

  ORA-00257: archiver error. Connect internal only, until this just released.

  To wear the DB back to business quickly, I tried to delete the old archive logs more than 1 day of the diskgroup first using the following command

  RMAN > delete noprompt archivelog until ' SYSDATE-1';

  So, I started to see the following message

  
  

  RECID = STAMP 19807 = 849396325 name=+ARCH_DG/bsoaprd/archivelog/2014_06_04/thread_2_seq_10066.331.849396309 archived log file

  Delete archived log

  RECID = STAMP 19809 = 849396940 name=+ARCH_DG/bsoaprd/archivelog/2014_06_04/thread_2_seq_10067.274.849396925 archived log file

  Delete archived log

  name=+ARCH_DG/bsoaprd/archivelog/2014_06_04/thread_2_seq_10068.341.849397529 RECID archived log file = 19810 STAMP = 849397543

  Delete archived log

  name=+ARCH_DG/bsoaprd/archivelog/2014_06_04/thread_2_seq_10069.298.849398167 RECID archived log file = 19812 STAMP = 849398182

  Delete archived log

  RECID = STAMP 19813 = 849398806 name=+ARCH_DG/bsoaprd/archivelog/2014_06_05/thread_2_seq_10070.317.849398783 archived log file

  Delete archived log

  name=+ARCH_DG/bsoaprd/archivelog/2014_06_05/thread_2_seq_10071.342.849399327 RECID archived log file = 19814 STAMP = 849399337

  Although I could see messages "deleted archived newspaper" as stated above there are 2 inconsistencies (odd behavior)

  1 inconsistency. Not a single byte released upwards of archive diskgroup. Each log archiving is 4 GB in size. I got "journal archived deleted" message at least 25 archived log files. For example, 100 GB should have been released.

  The diskgroup to archive has remained at 99.9% as shown below and the DB was inaccessible.

  SQL > select name,.

  Round(total_mb/1024,2) Total_GB.

  Round(free_mb/1024,2) Free_GB.

  free_mb/total_mb * 100 "% free"

  ABS ((free_mb/total_mb*100)-100) '% complete '.

  v $ asm_diskgroup;  2 3 4 5 6

  NAME TOTAL_GB FREE_GB % free full
  ------------------------------ ---------- ---------- ---------- ------------
  ARCH_DG 1000.78 99.9223633.077636719
  DATA_DG 10000.02 1188.19 11.8818616 88.1181384
  OCR_VOTE_DG 12 11.09 92.4623525 7.53764754
  < snipped >

  
  

  Inconsistency2. The number of available archived newspapers remained unchanged. He was always showing 301

  SQL > select count (*) from v$ archived_log where status = 'A ';

  COUNT (*)

  ----------

  301

  RMAN later > remove obsolete; command removed 17 archived newspapers.

  This means that 17 x 4 = 68 GB should have been released.

  But Diskgroup of free space and the County of v$ archived_log.status = 'A' is unchanged, as mentioned above.

  , Of ASMCMD, I deleted manually a full archive directory connects using rm-rf command.

  Even after running rm - rf on a whole directory full of archived newspaper space was not released and we always receive ""ORA-00257: archiver error '. "

  So, we had to shoot down the cluster using "stop crsctl crs. After the cluster has been raised using 'crsctl start crs' on both nodes,

  space has been freed up, and the DB became accessible.

  I wonder why the ASM diskgroup space was not release despite the removal of the archive logs. Are the 1 & 2 above mentioned inconsistencies caused by something handle to open the file ?

  Hello

  Please run command you at the level of the ASM, below

  SQL > alter diskgroup check all the repair;

  -At the end of the same thing, run below command and enable it to perform

  SQL > alter diskgroup rebalance power 4.

  -To validate the space

  $ asmcmd Pei lsdg

  Kind regards
  Loriette

• ASM Diskgroup

  Hello

  SQL > select NAME from v$ asm_diskgroup;

  NAME

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

  DATA

  FRA

  SQL > show disk parameter;

  VALUE OF TYPE NAME

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

  asm_diskgroups string FRA

  asm_diskstring string/dev/oracleasm/disks

  I wonder why these two commands produce two different values for asm_diskgroups?  Could someone please give me a hint?

  Hello

  The ASM Diskgroup can be started in multiple ways.

  11.2 docs

  On attachment to the boot disk groups

  At startup, the Oracle ASM instance try to mount the following disk groups:

  • Drive of the specified groups in the ASM_DISKGROUPS initialization parameter
  • Group of disks used by the Cluster Synchronization Services (CSS) for the vote of the files
  • Groups of disks used by Oracle Clusterware for Oracle Cluster Registry (OCR)
  • Group of disks used by the Oracle ASM instance to store the settings of the server ASM (SPFILE) file

  If no disk group are in the previous list, then the instance Oracle ASM do not mount at startup disk groups.

  What I see is happening... .i, please post below

  When you start the ASM, oracle will try read header of all the disk/LUN using asm_diskstring parameter in OCR/RSC (srvctl config asm) or spfile. If find ASM no matter what set of ASM DISKS it will try to mount automatically, because only when the DSO starting, he will deliver ALTER DISKGROUP MOUNT ALL.

  ASM find / know what are diskgroup can be mounted due head ASM DISK that contains this info on diskgroup.

  Other cases

  ASM is set on not (AUTO_START = never) automatically starts and is not not running, but the database is configured to start automatically (AUTO_START = always) and database dependency with diskgroup resource (for example, ora. DATA.dg), so during the startup of the crsd resource lance database dependency CSSD > ASM > DISGROUP > DATABASE

  11 GR 1 material and beyond... A lot of things (asm diskgroup, databases, etc.) became the resource managed by CRSD. Thus, by default, these resources are managed by CRSD not only by itself. I mean crsd knows and manage what resource needs to be started or a stop in the correct order.

  On your case a database resource can be begin your ASM diskgroup

  START_DEPENDENCIES = hard (ora. DATA.dg, ora. FRA.dg)

  This means that this resource will start only if these diskgroup and all its dependencies set first.

  Post edited by: Levi-Pereira

• ASM Diskgroup existing disks

  Hello
  Oracle RDBMS version 11.2.0.3.0
  OS: OEL 6.3
  Stand-alone server of the grid Infrastructure.
  MY question: I want to connect existing disks to ASM instance. What commands to create new disks existing diskgroup using without erasing the data on disks?

  Hello
  My recommendation is to create the ASM using ASCMA instance.

  Because the step below is valid only up to 11.1 or less, for 11.2 is missing a few extra steps.

  Now the steps for manual creation of DSO;

  1. create the file password for manually creating an ASM instance.
  orapwd file = $ORACLE_HOME/dbs/orapw + ASM Password123 = 5 entries = password

  In this way the user ASMSNMP will not be created and you must manually assigns a role SYSASM to the SYS user. ASMCA do automatically.

  2. create the needed directories:

  mkdir Pei $ORACLE_BASE/admin / + ASM
  CD $ORACLE_BASE/admin / + ASM
  mkdir bdump
  mkdir udump
  mkdir cdump
  mkdir pfile

  These directories are obsolete on 11.2, now Oracle use ADR to store logs.

  3. create the file init + ASM.ora under the $ORACLE_HOME/dbs:

  asm_diskgroups = 'DATA', 'FRA '.
  asm_diskstring = "/ dev/oracleasm/discs / *'"
  background_dump_dest ='/ u01/app/admin / + ASM/bdump "
  core_dump_dest ='/ u01/app/admin / + ASM/cdump '
  instance_type = "asm".
  LARGE_POOL_SIZE = 12 M
  user_dump_dest ='/ u01/app/admin / + ASM/udump '

  ASMCA value target memory at least 272 M

  4 - export ORACLE_SID = + ASM
  Instance ASM 5-start in editing mode:

  sqlplus / as sysasm
  Create spfile from pfile;
  Startup mount
  Show disc setting

  6 - Mount starts;

  change fra diskgroup Steed;
  change the data diskgroup medium;

  Missing a few steps here.

  ASMCA will activate/configure "ora.asm" (by adding the SPFILE and some PARAMETERS on OLR) resources and all dependent resources and create DISKGROUP resources.
  If you do not have config, ASM will be not started/stoped/restarted automatically.

  Once again... Is recommended to create an ASM Instance using ASMCA and a SMALL Lun (1 GB) to store only the ASM SPFILE when creating and then just get the other DISKGROUP.

  When you activate your database or add the DISKGROUP resources on Oracle Restart, and ASM must exist to work properly
  Published by: Levi Pereira April 11, 2013 13:31

• The ASM diskgroup in exadata space question

  All,
  
  I have an environment exadata 1 CCR node with 3 knots of strorage.
  
  My RECO ASM diskgroup is constantly showing the problems of space... Always higher than 80%.
  
  I just want to get a Board if I can remove mirroring and recover the space... Even though I know that the result of the deletion of the mirroring...
  
  Can I reduce the diskgroup RECO and create an another diskgroup?
  
  I want just a few ideas to do this without great risk involved...

  You may want to consider these MOS notes for more information on resizing the griddisks:
  Resize diskgroup without downtime to Exadata (Doc ID 1272569.1)
  How to resize the disk drive/grid ASM in environment Exadata (Doc ID 1245494.1)

  As others have mentioned, there is no way to change the level of redundancy in ASM, so you need to delete and re-create the diskgroup if you want to change to redundancy. We strongly discourage the use of EXTERNAL with Exadata redundancy.