ORA-27452 Error Creating Scheduler Job

When I create a job scheduler to perform a maintenance task gives me the following error.

ERROR at line 1:
ORA-27452: JOB _ / \ / \ / _ 10 is an invalid name for a database object.
ORA-06512: at “SYS.DBMS_ISCHED”, line 124
ORA-06512: at “SYS.DBMS_SCHEDULER”, line 271
ORA-06512: at line 2

In this case, the problem is in the string / \ / \ / (AN, my initials), have a habit of putting it to identify my things. Removing the “ANR” works. Another option is to use double quote in the definition of the job, as in the following example;

BEGIN
  DBMS_SCHEDULER.create_job (
    job_name        => '"JOB_/\/\_15"',
    job_type        => 'PLSQL_BLOCK',
    job_action      => 'BEGIN NULL; END;',
    start_date      => SYSTIMESTAMP,
    repeat_interval => 'freq=hourly; byminute=0; bysecond=0;',
    end_date        => NULL,
    enabled         => TRUE,
    comments        => 'Job de prueba /\/\/.');
END;
/

HTH – Antonio NAVARRO

Advertisements

ORA-29278 And ORA-24248 Error

After of upgrade a database, I have a job that send email with errors. Executing ad hod the procedure I see the following errors;

ORA-29278: error transitorio de SMTP: 421 Service not available
ORA-06512: en "SYS.UTL_SMTP", línea 54
ORA-06512: en "SYS.UTL_SMTP", línea 138
ORA-06512: en "SYS.UTL_SMTP", línea 699
ORA-06512: en "DBAOPER.SENDMAIL", línea 48
ORA-24248: seguridad extensible de XML DB no instalada
ORA-06512: en línea 3

In this case it is a 11g version. Search in MOS i see the note 602986.1, where It indicate that starting with 11g UTL_SMTP required XML DB (XDB), feature no installed in my database. The problem solve installing XDB.

HTH – Antonio NAVARRO

Be Careful With Unlimited Dump Traces

Yesterday I was on duty. When I was home I received a phone call from a work center operation (24×7) indicating that was filling a disk and there were a lot of alarms. After connect to the RAC and investigating the problem in more detail, watching the alert.log, alarms grid control, trace files, etc. I saw it had filled a filesystem that is assigned to the database. The problem was that someone had put a trace with size “unlimited” to a process. In this case the process is a demon or latent, which is always in machine, never ends. The trace was several days, which activate the trace don’t remember to put it off. The trace file has been growing, for gigabytes, to eat the whole filesystem. So when you use the next command;

alter session set max_dump_file_size = UNLIMITED ‘

Remember to be very careful.

HTH – Antonio NAVARRO

TCP Socket (KGAS)

This error is very typical especially in version 11g. Is a level problem sockets. From the side of the database we can not do anything because usually it come induced because of no response from  the client / service where it is called. In most of the cases is caused by calls to UTL_HTTP and UTL_TCP. Consider the following example.

select utl_http.request(‘http://sql1.wordpress.coN’) from dual

If we call a web page that does not exist, and we see from another session the event of waiting for the first session we will see this is the “TCP Socket (KGAS)”. On the other way, the first session, where we execute the select statement, will return the following error;

 REM The language is Spanish, but the error code are the same
select utl_http.request('http://sql1.wordpress.coN') from dual
       *
ERROR en línea 1:
ORA-29273: fallo de la solicitud HTTP
ORA-06512: en "SYS.UTL_HTTP", línea 1722
ORA-12535: TNS:timeout de la operación
ORA-06512: en línea 1

HTH – Antonio NAVARRO.

PRVF-7573 Sufficient Swap Size Is Not Available

Install a RAC on Solaris 10 box I get the next warning;

Swap Size - This is a prerequisite condition to test whether sufficient total swap space is available on the system.
  Check Failed on Nodes: [beer-2,  beer-1] 
Verification result of failed node: beer-2
Expected Value
 : 16GB (1.6777216E7KB)
Actual Value
 : 8GB (8388600.0KB)
 Details:
 - 
PRVF-7573 : Sufficient swap size is not available on node "beer-2" [Required = 16GB (1.6777216E7KB) ; Found = 8GB (8388600.0KB)]  - Cause:  The swap size found does not meet the minimum requirement.  - Action:  Increase swap size to at least meet the minimum swap space requirement.
Back to Top 
Verification result of failed node: beer-1
Expected Value
 : 16GB (1.6777216E7KB)
Actual Value
 : 16GB (1.6777208E7KB)
 Details:
 - 
PRVF-7573 : Sufficient swap size is not available on node "beer-1" [Required = 16GB (1.6777216E7KB) ; Found = 16GB (1.6777208E7KB)]  - Cause:  The swap size found does not meet the minimum requirement.  - Action:  Increase swap size to at least meet the minimum swap space requirement.

It is solved by increase swap area. In this case I have a device ready, you need a free device or add a new device;

swap -a /dev/dsk/pools/swap_2

HTH – Antonio NAVARRO

PRVF-4353 Maximum Open File Descriptors

Install a RAC in Solaris 10 I have received the next warning;

Soft Limit: maximum open file descriptors - This is a prerequisite condition to test whether the soft limit for "maximum open file descriptors" is set to at least 1024.
  Check Failed on Nodes: [beer-2] Check Succeeded On Nodes: [beer-1]  
Verification result of failed node: beer-2 
Expected Value
 : 1024
Actual Value
 : 256
 Details: 
 - 
PRVF-4353 : Proper soft limit for resource "maximum open file descriptors" not found on node "beer-2" [Expected = "1024" ; Found = "256"]  - Cause:  Soft limit for the resource does not meet the requirement on the specified node.  - Action:  Modify the resource limits to meet the requirement

The problem is a kernel parameter which is too low. You can change it including into the /etc/system the next line;

set rlim_fd_max = 1024

HTH – Antonio NAVARRO

Database hang

I reported an issue in a database. Looking at the grid control see the chart below.

Waits_for_Log_file_sync

Examining waits everything is staying in the “log file sync” event, this waiting says the log writer is writing the contents of the log buffer to redologs. Usually this is a problem of slow writing to disk. In this case it was a problem of SAN, once storage Team solves the issue, database returns to work fine.

HTH – Antonio NAVARRO