Recently, a customer reported that their oracle database could not be backed up, so I was surprised when I saw it remotely.
Rman tried to log in locally, and executed the command rman target / and found that it was indeed unable to log in, but the cause of the problem was very clear and very
It makes people laugh, but the hard disk space is full.
The specific error message screenshot is as follows:
According to the error message, it can be seen that the disk space is indeed exhausted, and the database instance cannot create the audit file. ORA-09945 is thrown, and the database is blocked.
rman initializes. Not only that, sqlplus / as sysdba also cannot log in to the database locally. The processing method is very simple, just try to
Log in to the server as root to enter the upper level of the oracle audit directory, the audit directory is not accessible, rename the audit directory mv, oracle
Execute a new audit directory, and then rm -rf forcibly delete it.
When database auditing is enabled, the audit directory will be used quickly, not only the disk space is easily exhausted, but also the inode resources of the file system.
It is easy to be exhausted. The reason is that the audit will generate many very fragmented and small files, which consumes the inode or disk space resources of the file system directory, resulting in
The database hangs. In the case of inode exhaustion, you need to expand the file system first, and then clean up small files. For the file system
Inode and space usage, it is best to set up monitoring in order to find and solve problems in time to prevent business interruption caused by database suspension.