Crash oraclehome node 1 - mistaken removal

Question

He was performing a procedure here to remove the spfile location to move storage, but control and came up with page break and deleted files that could not .. not delete them off cause I hit CTRL + C and stopped halfway exclusion.
The command was run this
rm -rf / u01/app/oracle/product/10.2.0
It was meant to be / u01/app/oracle/product/10.2.0/db1/dbs/spfileprod1.ora ...
I made a comparison of folders between nodes. I performed a binary copy of the node 2 to node 1, but was not done the "relink all ". realize the process of climbing the first instance and on it in stride, but when I connect via pl/sql or other application returns me the error ORA-12537 - TNS: connection closed
the database being run only by the node 2 and node 1 is shutdown.
What do I do to reinstall only the binaries of the database at node 1 and fix it? I measure by here we had no problem with that part of the clusterware
Database 10.2.0.5
Clusterware 11gr2
Red hat 5.6

Answer 1

user13521923 wrote:
He was performing a procedure here to remove the spfile location to move storage, but control and came up with page break and deleted files that could not .. not delete them off cause I hit CTRL + C and stopped halfway exclusion.
The command was run this
rm -rf / u01/app/oracle/product/10.2.0
First of all: WHY do you use the "-rf" option if your intention is only to delete a single file.
If you would have been more carefull it would have raised an error that your trying to delete a directory, which is not possible.
But anyway... none is perfect. You're lucky that you're on a RAC, otherwise your db would be completely down
It was meant to be / u01/app/oracle/product/10.2.0/db1/dbs/spfileprod1.ora ...
I made a comparison of folders between nodes. I performed a binary copy of the node 2 to node 1, but was not done the "relink all ". realize the process of climbing the first instance and on it in stride, but when I connect via pl/sql or other application returns me the error ORA-12537 - TNS: connection closed
the database being run only by the node 2 and node 1 is shutdown.
What do I do to reinstall only the binaries of the database at node 1 and fix it? I measure by here we had no problem with that part of the clusterware
Database 10.2.0.5
Clusterware 11gr2
Red hat 5.6First that comes into my mind is:
- How about the backup ?? Can you restore the installation directory from the backup ?
When restored, the node will probably start as normal
- If you're not able to restore the installation directory, then you'll need to dig into removing and re-adding nodes to your cluster. You'll probably end up here: http://download.oracle.com/docs/cd/E14072_01/rac.112/e10717/adddelclusterware.htm
Success!
FJFranken