Receiving ASM traces for errors: ORA-27090, LINUX-X86_64 ERROR: 17

 Increased audit files evidenced in the audit directory

CAUSE
The increased auditing is normal behaviour:
If your database generated 1 million files in 45 days it means that the ASM instance generated almost 2000 files every day

This is not so unusual for an ASM instance that has many rebalance operations or new disks added. The best thing to do in this situation is to create a purge job as per note 1298957.1.

With this in place the audit files will not accumulate .

SOLUTION
 1.  The increased auditing is not the reason for the error:  ORA-27090: Unable to reserve kernel resources for asynchronous disk I/O
Linux-x86_64 Error: 17: File exists

The auditing is normal for an ASM instance with heavy activity.

2.  However the ORA-27090 error does indicate that you are running into an OS resourcing issue.

The number of aio requests (aio-nr) is likely high during this error  (in this case 1661256).

And may have reached the aio-max-nr value.  Please check with your system administrator

To determine the cause of the high number of current system-wide number of asynchronous I/O

requests seen.

3.  Example of aio values:

[oracle@test ~]$ cat /proc/sys/fs/aio-nr
  1661256
  [oracle@test ~]$ cat /proc/sys/fs/aio-max-nr
  3145728