UTIL_File In Remote Node

Hi,
My Problem is that we seperated our AppServer(Solaris) and DBServer(Solaris) and some of my Script uses UTIL_FILE to write file from database,so i mounted My Appserver in DBServer so that it writes the data in Appserver only,now my Mount is working fine still i'm getting Oracle Error
ORA-29283: invalid file operation
ORA-06512: at "SYS.UTL_FILE", line 488
ORA-29283: invalid file operation
And i have given the Folder Full permission(777) but still no use,
so is that a permission issue or My SP is not working

Hi,
has the user already assigned Write Privilege to the Virtual Directory.
777 is in the OS Level, but in the DB Level, SYS need to grant privilege to the user.
HTH

Similar Messages

The selected remote nodes 'node3' are not accessible

Hello,
I removed node3 from 2 node RAC Cluster.
After this I have re-add this node to my RAC. I added without any problem to CRS but I have a problem with script addNode.sh (/u01/app/oracle/product/crs/oui/bin).
I have to separate homes for ORACLE_HOME and ASM_HOME, so I need to run this script twice for each ORACLE_HOME.
I run addNode.sh from one of the cluster nodes (node1), if I run:
export ORACLE_HOME=/u01/app/oracle/product/11.1.0/asm
cd $ORACLE_HOME/oui/bin
./addNode.sh -silent CLUSTER_NEW_NODES={node3} CLUSTER_NEW_PRIVATE_NODE_NAMES={node3-priv} CLUSTER_NEW_VIRTUAL_HOSTNAMES={node3-vip}
I got the following error:
Performing tests to see whether nodes node2,node3 are available
............................................................SEVERE:The selected remote nodes 'node3' are not accessible.
... 100% Done.
I reviewed logs located in /u01/app/oraInventory/logs I found
Status of node 'node3':
null
The virtual host name 'node3-vip' for node 'node3' is already in use.
I checked crs_stat from $ORA_CRS_HOME, but all applications (ons, gsd, vip) and targets are ONLINE. I tried to run vipca -nodelist node1,node2,node3 and updated a configuration. I checked /etc/hosts, connection via SSH between all hosts belonging to RAC: node1, node1-vip, node1-priv etc. Everything looks like fine, so I have no idea what caused this problem.
Software which I used:
Linux: UNL 5.4 x86_64
11g Release 11.1.0.7.0 - 64bit, Oracle Clusterware
Thank you for your answer,
Best Regards,
Edited by: user4924714 on 2010-02-15 07:50

Hi,
Thanks for your reply.
I am running addNode.sh from node which already exist in the Rac Cluster. I tried on node1, node2 but the resulat are the same.
Best regards,

11gr2 - 11.2.0.2 -root.sh executed on remote node first can harm anything ?

Hello Expert,
It seems like we have Oracle GRID forum having some issue so i am posting grid Question on Real Application Clusters forum.
I am installing 11gr2 - 11.2.0.2 Grid infrastructure on Red hat linux 5- x86_64 (Host01-04, Host01-LOCAL and Host 02-04 REMOTE) . During install it pop up to execute
orainstRoot.sh,root.sh as root and they clearly mentioned that execute
orainstRoot.sh,root.sh on local node first and then the remote node. But by mistake i have executed root.sh on remote node first HOST02 (Host01-04, Host01-LOCAL and Host 02-04 REMOTE).
Does it really affect anything?
Thanks in Advance.
Billy

Yes, you're right.
Its a empty Prod Box where nothing is running but eventually it would become Production LIVE box with Prod Database instances.
Yes, Only concerns about the master instance .As you seen below host 2 is master right node right now.without bringing down the node 2 can i convert node 1 to master.
we are not using ASM so you'll see ASM services in OFFLINE state.
HOST1: crsstat
Resource Target State
ora.LISTENER.lsnr ONLINE ONLINE on host1
ora.LISTENER_SCAN1.lsnr ONLINE ONLINE on host1
ora.LISTENER_SCAN2.lsnr ONLINE ONLINE on host3
ora.LISTENER_SCAN3.lsnr ONLINE ONLINE on host2
ora.asm OFFLINE OFFLINE
ora.cvu ONLINE ONLINE on host2---
ora.gsd OFFLINE OFFLINE
ora.net1.network ONLINE ONLINE on host1
ora.oc4j ONLINE ONLINE on host2---
ora.ons ONLINE ONLINE on host1
ora.registry.acfs OFFLINE OFFLINE
ora.scan1.vip ONLINE ONLINE on host1
ora.scan2.vip ONLINE ONLINE on host3
ora.scan3.vip ONLINE ONLINE on host2
ora.host1.ASM2.asm OFFLINE OFFLINE
ora.host1.LISTENER_host1.lsnr ONLINE ONLINE on host1
ora.host1.gsd OFFLINE OFFLINE
ora.host1.ons ONLINE ONLINE on host1
ora.host1.vip ONLINE ONLINE on host1
ora.host2.ASM1.asm OFFLINE OFFLINE
ora.host2.LISTENER_host2.lsnr ONLINE ONLINE on host2
ora.host2.gsd OFFLINE OFFLINE
ora.host2.ons ONLINE ONLINE on host2
ora.host2.vip ONLINE ONLINE on host2
ora.host3.ASM3.asm OFFLINE OFFLINE
ora.host3.LISTENER_host3.lsnr ONLINE ONLINE on host3
ora.host3.gsd OFFLINE OFFLINE
ora.host3.ons ONLINE ONLINE on host3
ora.host3.vip ONLINE ONLINE on host3
ora.host4.ASM4.asm OFFLINE OFFLINE
ora.host4.LISTENER_host4.lsnr ONLINE ONLINE on host4
ora.host4.gsd OFFLINE OFFLINE
ora.host4.ons ONLINE ONLINE on host4
ora.host4.vip ONLINE ONLINE on host4
Billy.

Want to keep the data on one node and allow remote nodes to read it?

by default, we have a distributed cache with backupcount=1 and data is kept over all the members in equal parts
is there a way to do set it another way: i need only one 'main' node to put entries into the cache and keep them locally, while remote nodes can only read entries? the reason for that - i want to update the entries on the main node quite frequently (do not them travel anywhere over the network) - while remote read operations are infrequent. Also, Iwant the whole dataset survive the shutdown of any secondary node. In other words, I need 'all eggs in one backet' with remote read access.

Andrey -
Provided the 'remote/secondary' nodes are within the local area network, you could use a dedicated distributed cache service; storage enabled
on the 'main' node and storage disabled on the 'secondary' nodes. Access from outside the LAN will require configuring a proxy service on the
'main' node and 'secondary' node access over Coherence*Extend.
/Mark

Integration Broker - PTool 8.50 - Pinging remote node errors out

Hi! I am trying to configure default local nodes on HCM 9.1 and FSCM 9.1, both systems have PeopleTools 8.50.
HCM 9.1 is installed on Windows 7 Home Premium. DB is Oracle 10g
FSCM 9.1 is installed on Windows 2003 server. DB is Oracle 10g
Problem:
=====
When I ping remote node from HCM 9.1 it displayes the following message
“PeopleSoftTargetConnector - The application server is down (or not responding). Check AppServer URL.”
When I ping the remote node from FSCM 9.1 then I get the following message:
“PSFT Authentication token failed for Node PSFT_HR. (158,454)”
Setup Details:
=========
Here are steps I followed on both systems:
FSCM 9.1. side :
=========
1.Defined local gateway. http://<My-Machine B>:/PSIGW/PeopleSoftListeningConnector When I ping it from Gateway page, it shows it is active.
2.Renamed default node PSFT_EP to PSFT_TRN. Before renaming ran the appmsgpurgeall.dms script from <PS_HOME>\scripts\ to make sure there are no references to PSFT_EP in runtime tables. This is PS supplied dms script and they recommend to run this script before renaming any node (It should be done at the initial setup time).
3.On the Node page, PSFT_TRN node is defined as default node. Specified the default user Id as ‘PTINT’ the common User id created by cloning ‘VP1’. On the ‘Connectors’ tab Gateway ID is ‘LOCAL’ and connector id is ‘PSFTTARGET’. From the ‘Connectors’ tab Pinged the PSFT_TRN node and it was pinged successfully.
4.On Gateway Properties page, defined "APP Server URL" as ‘<Machine Name B>:9000, user Id/pwd as ‘PTINT’ and tool release as ‘8.50’.
5.Defined PSFT_TRN node, with APP Server URL as ‘<Machine Name B>:9000, user Id/pwd as ‘PTINT’ and tool release as ‘8.50’. When I ping this node, it is pinged successfully.
6.Inserted another node PSFT_HR, with APP Server URL as ‘<Machine Name A>:9000, user Id/pwd as ‘PTINT’ and tool release as ‘8.50’. When I ping PSFT_HR node from this page I get the following error:
“PSFT Authentication token failed for Node PSFT_HR. (158,454)”
HCM 9.1. side:
=========
1.Defined local gateway. http://<My-Machine A>:/PSIGW/PeopleSoftListeningConnector. When I ping it from Gateway page, it shows it is active.
2.On the Node page made sure ‘PSFT_HR’ node is default and active. Specified the default user Id as ‘PTINT’ the common User id created by cloning ‘PS’. On the ‘Connectors’ tab Gateway ID is ‘LOCAL’ and connector id is ‘PSFTTARGET’. From the ‘Connectors’ tab clicked on ‘Ping Node’ button for PSFT_HR node and it was pinged successfully.
3.On Gateway Properties page, defined "APP Server URL" as '<Machine Name A>:9000', user Id/pwd as ‘PTINT’ and tool release as ‘8.50’.
4.Defined PSFT_TRN node, with "APP Server URL" as ‘<Machine Name B>:9000', user Id/pwd as ‘PTINT’ and tool release as ‘8.50’. When I ping this node, the following message is displayed:
“PeopleSoftTargetConnector - The application server is down (or not responding). Check AppServer URL.”
Basically, I am unable to ping remote node on both systems. Local node ping successfully on both systems.
Has any one encountered similar problem and found a solution?
Thanks
Raj
Edited by: 865522 on Jun 13, 2011 2:18 PM

Yes, I could ping both machines from each other.
Here is the host file contents from Windows 7:
127.0.0.1 localhost
::1 localhost
127.0.0.1 activate.adobe.com
192.168.1.106 RAJ-PC
Here is from Window 2003 server:
127.0.0.1 localhost
192.168.1.100 saroja
Both nodes are listed under PeopleTools>Security Objects> Single Signon
Both nodes are secured with userid/pwd on Node page.
One of my machine had fire wall on, so I swtiched the fire wall off. Now machine A can ping both nodes but machine B can ping only local node. Pinging remote node displays "User name not defined in database".
Raj
Edited by: Raj on Jun 13, 2011 11:37 PM

How to asmcmd cp non ASM files to ASM on a remote node

We have a 8TB database, RDBMS 11,1. on Solaris 10, non ASM. Need to copy the 1290 files to remote node of AIX ASM to prepare for a upgrading to 11.2.0.2.
We tried to use asmcmd cp to copy the files directly from non ASM too ASM remotely, but did not work.
$ asmcmd cp -ifr [email protected]:/oradata/prdsdev/datafiles/dev_4.dbf +PRDSU_DATA/PRDSU/TEMPFILE/dev_4.dbf
Enter password: ********
Argument "normal" isn't numeric in numeric ge (>=) at /optware/oracle/11.2.0.2/grid/lib/asmcmdshare.pm line 1676.
errors
ORA-01031: insufficient privileges (DBD ERROR: OCISessionBegin)
We were using the user fttr. Do we need to login as SYSASM, do we need to go through OS auth, any other privilege required. What instance do we need to login, regular database that is using the data files, or ASM instance is goin g to manage the files.
Thanks in advance

1. my previous question is if asmcmd cp can accomplish the task.
2. RMAN COPY, as I know, will either copy non asm files to ASM files locally, or copy files asm to asm across nodes. which means that we have to do two steps, rather copy the non asm files to a node of asm directly.

DBCA does not show the remote nodes on a RAC 10.2.0.2 HP-ITANIUM

HI at all,
I have a problem and I hope someone can help me.
A collegue have installed CRS and the Oracle HOME for a RAC 10.2.0.2 on a HP-ITANIUM ( two nodes) .
Im'm using the dbca for creating a cluster database.
I can see the RAC option pages , but in the "select nodes page" I found only the local node and not the remote node.
Any suggestion is appreciated
thank you
Adriano
$ crs_stat
NAME=ora.itmicz50.LISTENER_ITMICZ50.lsnr
TYPE=application
TARGET=ONLINE
STATE=ONLINE on itmicz50
NAME=ora.itmicz50.gsd
TYPE=application
TARGET=ONLINE
STATE=ONLINE on itmicz50
NAME=ora.itmicz50.ons
TYPE=application
TARGET=ONLINE
STATE=ONLINE on itmicz50
NAME=ora.itmicz50.vip
TYPE=application
TARGET=ONLINE
STATE=ONLINE on itmicz50
NAME=ora.itmicz51.LISTENER_ITMICZ51.lsnr
TYPE=application
TARGET=ONLINE
STATE=ONLINE on itmicz51
NAME=ora.itmicz51.gsd
TYPE=application
TARGET=ONLINE
STATE=ONLINE on itmicz51
NAME=ora.itmicz51.ons
TYPE=application
TARGET=ONLINE
STATE=ONLINE on itmicz51
NAME=ora.itmicz51.vip
TYPE=application
TARGET=ONLINE
STATE=ONLINE on itmicz51
and the olsnodes command :
$ olsnodes
itmicz50
itmicz51
Message was edited by:
user549224
null

HI Chandra,
thank you very much for your reply. I suppose the oraInventory is the same, however the contents of inventory.xml file is this:
<?xml version="1.0" standalone="yes" ?>


<INVENTORY>
<VERSION_INFO>
<SAVED_WITH>10.2.0.1.0</SAVED_WITH>
<MINIMUM_VER>2.1.0.6.0</MINIMUM_VER>
</VERSION_INFO>
<HOME_LIST>
<HOME NAME="OUIHome1" LOC="/work/app/oracle/product/CRS" TYPE="O" IDX="1" CRS="true">
<NODE_LIST>
<NODE NAME="itmicz50"/>
<NODE NAME="itmicz51"/>
</NODE_LIST>
</HOME>
<HOME NAME="OUIHome2" LOC="/work/app/oracle/product/RAC10g" TYPE="O" IDX="2">
<NODE_LIST>
<NODE NAME="itmicz50"/> <===== EVIDENCE
</NODE_LIST>
</HOME>
</HOME_LIST>
</INVENTORY>
I can see that within the node_list tag I have only one hostname It is correct?
thank you in advance
Adriano Capruzzi

Unabe to ping Remote Node

Hi:
I have PS FMS 9.0 and PS HRMS 9.0 IBs (both PT 8.49.18) configured for SSO and messaging. I am able to ping remote node - PSFT_EP from HRMS environment but unable to ping remote node - PSFT_HR from FMS environment. I have performed the steps mentioned in the following thread:
Re: PeopleSoft Integration Broker from FSCM9 [PT8.48] to CRM9 [PT8.49]
But still I get the following message in app server log file:
PeopleSoft Token authentication failed: token has expired: VP1@JavaClient
Any suggestions?
Thanks,
Sameer.

Is your gateway configured on SSL?
If its the default value on both the sides is same you could try the following:
1> Ensure the time zone diff (*IF ANY*) b/w the web n App server of the HR & FMS is catered for.
2> Try increasing the log fence to 5 on both the sides and ping the nodes to capture the Auth Token value.

SSH User equivalence is not set up on the remote nodes

I'm getting the above message when trying to push deploy a Linux agent 10.2.0.2 from an OEM.
I have run the sshUserSetupNT.sh succesfully - I believe there were no error messages being returned.
I can cygwin ssh to connect successfully to the linux box and issue commands without providing a password.
- and I've run the failing command from the log below successfully from a command prompt. (changed it to run date)
C:/cygwin/bin/ssh.exe -o FallBackToRsh=no -o PasswordAuthentication=no -o StrictHostKeyChecking=yes -l oraoem centos5 /bin/sh -c '/bin/date '
I appears to be complaining about an RSA key - but how can that be when I can connect from a command prompt succesfully.
Any suggestions!
snip from deployment logfile
=======================================================================
Performing check for CheckSSHEquivalence
Does SSH equivalence exist with the remote nodes
Setting username to oraoem
Setting FallbackToRSH to false
Setting REMOTE_PATH_PROPERTIES_LOC to D:\OracleHomes\oms10g\/sysman/agent_download/10.2.0.2.0/linux/agentdeploy
C:\WINDOWS\system32\ping.exe#centos5#-n#5#-w#5000
C:\WINDOWS\system32\ping.exe centos5 -n 5 -w 5000
Exitcode of the remote command [rnode centos5]: 0
Errstream [rnode centos5]:
CommandResult status [rnode centos5]: true
C:/cygwin/bin/ssh.exe#-o#FallBackToRsh=no#-o#PasswordAuthentication=no#-o#StrictHostKeyChecking=yes#-l#oraoem#centos5#/bin/sh#-c#'/bin/true '
C:/cygwin/bin/ssh.exe -o FallBackToRsh=no -o PasswordAuthentication=no -o StrictHostKeyChecking=yes -l oraoem centos5 /bin/sh -c '/bin/true '
Exitcode of the remote command [rnode centos5]: 255
Errstream [rnode centos5]: No RSA host key is known for centos5 and you have requested strict checking.
Host key verification failed.
CommandResult status [rnode centos5]: false
Expected result: centos5
Actual Result: Problem: 'PROV-16018: The command C:/cygwin/bin/ssh.exe failed.'
Check complete. The overall result of this check is: Failed <<<<
Check complete: Failed <<<<
Problem: SSH equivalence check has failed
Recommendation: Setup SSH equivalence with the remote hosts and retry

Reboot the box and try again.

Restart adminserver from a remote node?

Greetings,
Is there a command to restart ( not stop / start ) admin server from a remote node using nodemanager?
thanks
Seetharaman

Hi Seetharaman,
I don't think so you can restart the admin server using nodemanager without stop and start as there is no dedicated command for it. The commands with Node Manager Commands can be seen below.
Topic: Node Manager Commands
http://download.oracle.com/docs/cd/E13222_01/wls/docs90/config_scripting/reference.html#1030876
Regards,
Ravish Mody

Installing RAC Problem: Oracle Clusterware 10g is present on remote node

Folks,
Hello.
I am installing Oracle 11gR2 RAC using 2 Virtual Machines (rac1 and rac2 whose OS are Oracle Linux 5.6) in VMPlayer.
I have been installing Grid Infrastructure using runInstaller in the first VM rac1 from step 1 to step 9 of 10.
On the step 9 of 10 in the Wizard, accidentally, I touch the Mouse, and the Wizard is gone.
The directory for installing Grid in the 2 VMs is the same: /u01
In order to make sure everything is correct, I delete entire directory /u01 in the 2 VMs and install Grid in rac1 again.
But get the error message as follows:
[INS-40915] The Installer has detected the presence of Oracle Clusterware 10g on the remote node rac2.localdomain.
Cause: Oracle Clusterware 10g is present on remote node rac2.localdomain but not on the local node.
Action: To upgrade the Oracle Clusterware 10g Installation to this release, run Installer on a node where Oracle Clusterware 10 installation is located. To create a new cluster, choose a set of nodes that are not part of an existing cluster.
Regarding the above error message, I do not install any Oracle Clusterware 10g in rac2 because it's the first time when I install Grid using rac1 and rac2.
My question is:
Do any folk understand how to remove something in rac2 to solve the above issue so that I can install Grid again ?
Thanks.

Folks,
Hello. Thanks a lot for replying.
The tutorial http://docs.oracle.com/cd/E11882_01/install.112/e22489/rem_orcl.htm#CBHEFHAC is about the regular procedure to deinstall Grid Clusterware and Database.
I have just installed Grid and have not installed the Database. I have removed ORACLE HOME /u01 completely in both VMs. There is no way to follow the regular procedure to deinstall at this time.
I do not install any Oracle Clusterware 10g in the VM rac2 because it's the first time when I install Grid using VM rac1 and rac2.
When install Grid again. I get the error message:
"[INS-40915] The Installer has detected the presence of Oracle Clusterware 10g on the remote node rac2.localdomain."
This is strange! It seems that we need to remove some files in the VM rac2 like /etc/oracle that I am not sure.
My questions are:
First, where is 10g directory to remove ?
Second, what files left on the system to clean up ?
Third, what does " Follow 239998.1" means regarding asanqa's replying to me ?
Thanks.

ORA-15032 when mounting a diskgroup in a remote node

Hi all,
Oracle 10g R2 (10.2.0.4)
Windows 2003 SP2 Server
I successfully went through the steps of creating an Oracle RAC of two nodes rac1 and rac2 till I reached to creating ASM step.
To create an ASM, I issused dbca from rac1 and created an ASM instance. The issue appeared when I created the Diskgroup.
After selecting the disk and clicking on OK, I got the following error:
http://www.ahmedbaraka.com/temp/errorsnapshot.png
Following are the generated log files in rac1:
C:\oracle\product\10.2.0\db_1\cfgtoollogs\dbca\trace.log http://www.ahmedbaraka.com/temp/trace.log
C:\oracle\product\10.2.0\db_1\log\rac1\racg\ora.rac1.ASM1.asm.log http://www.ahmedbaraka.com/temp/ora.rac1.ASM1.asm.log
Following are the generated log files in rac2:
C:\oracle\product\10.2.0\admin\+ASM\bdump\alert_+asm2.log http://www.ahmedbaraka.com/temp/alertasm2.log
Any hint is appreciated as I'm stuck with this for a couple of days.

Hi Philippe Florent,
You were right about the version, so I applied the patch and redid the procedure. Unfortunately, I still face exactly the same error.
I also made sure when I make changes on a disk in rac1 (in Windows level), it is automatically seen by rac2.
Below are links of the new logs and pfiles:
http://www.ahmedbaraka.com/temp/alertasm22.log
http://www.ahmedbaraka.com/temp/trace2.log
http://www.ahmedbaraka.com/temp/rac1pfile.ora
http://www.ahmedbaraka.com/temp/rac2pfile.ora
Output of the commands you asked about:
# RAC1
SQL> select name, state from v$asm_disk ;
NAME                           STATE
                               NORMAL
DGDATA1_0000                   NORMAL
# RAC2
1* select name, state from v$asm_disk
SQL> /
NAME                           STATE
                               NORMAL
                               NORMALThanks in advance.

Error connection with a database in remote node

When I try to start a dialog instance does not detect that the database is in a remote server is booted:
"Checking SAP R/3 PAP Database
Database is not available via /usr/sap/PAP/SYS/exe/run/R3trans -d -w
Database PAP must be started on remote server"
The output of R3trans -x is:
"This is ./R3trans version 6.05 (release 46D - 28.03.01 - 11:30:00).
2EETW169 no connect possible: "DBMS = ORACLE      --- dbs_ora_tnsname = 'PAP'"
./R3trans finished (0012)."
The output of trans.log is:
"4 ETW000 [developertrace,0] -->oci_logon(con_hdl=0, user='', dbname='PAP')
4 ETW000                                                         88 0.035395
4 ETW000 [dboci.c ,0] *** ERROR => OCI-call 'olog' failed: rc = 12705
4 ETW000                                                      75859 0.111254
4 ETW000 [dbsloci.,0] *** ERROR => CONNECT failed with sql error '12705'"
Can anybody help me to solve this problem?

Thank you for your help, RK.
The problem was that sidadm was not the owner of ORA_NLS.
Now I have another problem: when running startsap not boot disp + work.

Running scandisks command in remote nodes

Grid Infrstructure version: 11.2.0.4
Platform :Oracle Linux 6.4
We have a 3-Node production RAC cluster . We have created a 3-Node test RAC cluster for cloning purposes
We are automating the SAN level cloning from production to test. After the Hitachi guys take a snapshot of diskgroup LUNs from production they present the LUNs to the test RAC cluster. At test the cluster we need to do the following to do the cloning
Step1. In Node1, run
/etc/init.d/oracleasm scandisks
Step2. In Node2 run
/etc/init.d/oracleasm scandisks
Step3. In Node3 , run
/etc/init.d/oracleasm scandisks
Step4.
srvctl start diskgroup -g <dg_name> -n Node1
srvctl start diskgroup -g <dg_name> -n Node2
srvctl start diskgroup -g <dg_name> -n Node3
Step5. Finally, Start the database start the database in test cluster using
srvctl start database -d MANHPRD
From a shell script, I would like to run the above Step1,2 and 3 from Node1 in a sequential order as shown above . How can I do this ? I am not worried about Step4 and 5 as it can be run from any nodes.

Hello Page,
why not use ssh? very simple
from node 1:
ssh root@node2 /etc/init.d/oracleasm scandisks
take note of Dude! comment
edited: take a look at Tobi&#39;s Oracle DBA &amp; UNIX Blog: Cloning ASM Diskgroup to Create Database Clone

Error starting observer from remote node

Hello all.
I have a two nodes Veritas (not important) cluster with Data Guard. I don't have shared drives, I know!
Now, I had configured an observer in one of the nodes which was working fine but for obvious reasons, I tried to move it to a third node.
This is the error I'm getting and I have no idea how to solve it out. What surprises me is that the wrong "parameter" is not shown even in Metalink.
{code}
DGMGRL> connect sys@primary
Password:
Connected.
DGMGRL> show configuration
Configuration - testing-replication
Protection Mode: MaxAvailability
Databases:
    primary - Primary database
      Warning: ORA-16819: fast-start failover observer not started
    standby - (*) Physical standby database
      Warning: ORA-16819: fast-start failover observer not started
Fast-Start Failover: ENABLED
Configuration Status:
WARNING
DGMGRL> start observer;
Corrupted document
<FAST_START_FAILOVER enabled="yes" MIV="24" OBID="553124169" threshold="15" obsv_recn="0" auto-reinstate="yes"><PRIMARY name="primary">primary</PRIMARY><STANDBY name="standby">standby</STANDBY></FAST_START_FAILOVER>
Error: LPX-00106: Warning: attribute "obsv_recn" of element "FAST_START_FAILOVER" is undefined
Error: ORA-16591: unknown field "obsv_recn" in document
Error: ORA-16593: XML conversion failed
Failed.
{code}
Thanks for any help you can give me.
Alex.

Hmm. Well, I don't see how would that help. It's the same command I'm running but in silent mode. The result is the same.
What bothers me is the message:
Error: ORA-16591: unknown field "obsv_recn" in document
I couldn't find anything related to that field/parameter on the web. Also, what's the "document" is referring too? fsfo.dat? In that case, is blank in my case.
Again, I've set this up on another server (one of the nodes where the databases resides) and created the configuration "create configuration" and "enable configuration" from there, but not here.
Do I have to remove the configuration and start all over again from this third node?
Thanks,
Alex.

UTIL_File In Remote Node

Similar Messages

Maybe you are looking for