Can a dbms_job.run execute a job that execute another dbms_job.run?

Hi,
I'm using Oracle9iR2 database.
I create a package procedure that loop through a list of jobs from user_jobs table and run it using dbms_job.run(jobid).
Then I schedule this package procedure as a job (say JobA) itself to execute the above procedure every 15 minutes.
However, it always fail. When I execute JobA manually from sqlplus, I get the folllowing error. Any idea what's wrong?
SQL> exec dbms_job.run(55805);
BEGIN dbms_job.run(55805); END;
ERROR at line 1:
ORA-12011: execution of 1 jobs failed
ORA-06512: at "SYS.DBMS_IJOB", line 406
ORA-06512: at "SYS.DBMS_JOB", line 272
ORA-06512: at line 1
Please advise.
Thank you.
Message was edited by:
bchurn

Hi,
I am bchurn's colleague who is currently also looking into this problem. The code that we have is similiar to yours except there's a cursor looping through all the failed jobs and executing them.
I am not sure what is the actual difference but it is actually a job running a procedure instead of through SQL Plus.
However, we found out that a job cannot run another job. But it seems to be alright with your settings :-/
What we have is actually a job created for every email to be sent out to our users (jobA). But many a times, the emails were not sent out because of unknown problems (during that time - I think it's the ORA-00600 error).
The last time we created the second job (jobB) for the purpose of re-executing failed jobA jobs. The errors are that a job cannot execute another job.
Then, we worked around to change jobB to re-adjust the execution time for all the jobA jobs instead of directly reexecuting them. While some failed jobA jobs are executed successfully because of it, some still remains. We found out that the errors from failed execution of jobA jobs as following:
ORA-00600: internal error code, arguments: [kgassg_2], [], [], [], [], [], [], []
ORA-29279: SMTP permanent error: 555 5.5.4 <fieldNameHere> parameter unrecognized
The second error very seldom. First one most common. The jobs eventually being executed successfully after many re-execution (sometimes about 3 to 4 times, some over 20 or 60 times). Anyone can give suggestions or inputs on the first error? The error log file and its trace files do not give much information on the errors with something like the following:
Error log file:
Errors in file /opt/app/oracle/products/920/admin/testapp/bdump/somefilename.trc:
ORA-00600: internal error code, arguments: [kgassg_2], [], [], [], [], [], [], []
And when I go and search for the somefilename.trc, it does not exist.
Any idea on how to solve this or identifying the actual problem?
Thank you in advance.

Similar Messages

HT202696 Can I renew my Applecare Protection Plan that is due to run out?

Can I renew my Applecare Protection Plan that is due to run out?

Applecare is designed as an extended warranty for a set period (beyond the original manufacturer's warranty). It is a one time deal and not at all intended as an open ended insurance policy. If you have issues, you can still contact apple, but may have to pay for the call. You can post issues here. Visit an Apple store genius bar or authorized service center (which may involve charges for an services performed). Basically, it is what would happen for any electronic device that is out of warranty.

How do I create an executable with arguments that executes once and terminates

I'm just starting with LabView, so I'm sure this issue has been address somewhere already. I just can't locate the answer.
I am running version 6i (I may be able to get 7 or 8).
I need to create a command line executable program with parameter input that executes only once and then terminates.
This is for instrument control (sets the signal generator frequency, level, RF on/off and arms the Trigger). I have written a simple VI that works from within LabView (programs the instrument when the Run button is pressed). I have further managed to perform an Application Build to create an executable however, the application continues to run, waiting for the Run button to be pressed. Also, I need to know how to pass parameters to this executable (freq, level, RF on/off, trigger arm).
This executable will (eventually) be called repeatedly by a Matlab file.
Thanks,
D

LabVIEW 7 added a property to get a command line argument. For older versions see http://sine.ni.com/apps/we/niepd_web_display.display_epd4?p_guid=B45EACE3DA5B56A4E034080020E74861.
I'm not sure I understand you problem with the Run button. If you set the execution mode to Run When Opened from VI Properties, it should execute once whenever it's called. I think that when building an executable, it changes the property to this if it's not already set. If the application does not have this set, it will not be running at all. I'm assuming that you are talking about the Run button on the toolbar and not a run Boolean that you've created on the front panel.
Since you have the app builder, your other option is to create a DLL. If you can call a DLL in Matlab, then that would probably be simpler.
Message Edited by Dennis Knutson on 11-28-2005 02:46 PM

User Executes Stored Procedure That Executes sp_send_dbmail; receives email but also gets error Cannot alter the queue 'ExternalMailQueue', because it does not exist or you do not have permission.

Using SQL Server 2012 SP1
I have a user that is submitting a procedure that uses sp_send_dbmail. I have also noticed that they have the following code in their procedure
DECLARE @rc INT
IF NOT EXISTS (SELECT * FROM msdb.sys.service_queues
WHERE name = N'ExternalMailQueue' AND is_receive_enabled = 1)
EXEC @rc = msdb.dbo.sysmail_start_sp
The user submits the procedure and she gets the email but she also gets the following error message:
Msg 15151, Level 16, State 1, Procedure sysmail_start_sp, Line 8
Cannot alter the queue 'ExternalMailQueue', because it does not exist or you do not have permission.
Mail (Id: 2402) queued.
(1 row(s) affected)
I have granted execute to sp_send_dbmail and sysmail_start_sp. I have also granted select to service_queues.
Does anyone have any solutions for the above error message?
lcerni

The contents of sysmail_start_sp is this:
CREATE PROCEDURE sysmail_start_sp
AS
SET NOCOUNT ON
DECLARE @rc INT
   DECLARE @localmessage nvarchar(255)
ALTER QUEUE ExternalMailQueue WITH STATUS = ON
SELECT @rc = @@ERROR
IF(@rc = 0)
BEGIN
ALTER QUEUE ExternalMailQueue WITH ACTIVATION (STATUS = ON);
   SET @localmessage = FORMATMESSAGE(14639, SUSER_SNAME())
   exec msdb.dbo.sysmail_logmailevent_sp @event_type=1, @description=@localmessage
END
RETURN @rc
The user get the error, because she does not have any permission on the queue in question. To be able to alter the queue, the following applies according to Books Online:
Permission for altering a queue defaults to the owner of the queue, members of the db_ddladmin or db_owner fixed database roles, and members of the sysadmin fixed server role.
Note that is would be db_ddladmin or db_owner in msdb. Now, supposedly the queue is already active, and in that case I think that it is sufficient that the user has VIEW DEFINITION on the view. This would permit her to see the row in sys.service_queues,
why there would be no need to call sysmail_start_sp.
Altertanively, change the check to:
DECLARE @isenabled bit
SELECT @isenabled = (SELECT is_receive_enabled FROM msdb.sys.service_queues
               WHERE name = N'ExternalMailQueue')
IF @isenabled = 0
   EXEC @rc = msdb.dbo.sysmail_start_sp
The idea here is that, if the user has no permission to read the information in the DMV, @isenabled will be NULL, and you just pray and hope that the queue is up and running.
I suspect that the reason this worked for you on SQL 2005 is that on that instance someone at some point in made all the required configurations for it to work, but all that is forgotten now. That happens to me too.
If you really want the user to be able to start the queue, I have some better ideas than adding her to a role, but I skip the details for now.
Erland Sommarskog, SQL Server MVP, [email protected]

Using Layout Variant in jobs that are scheduled to run in background

Hi Guys,
I have an issue in using the "layout variant" in running a report in the background. Here is what i do to run a report.
I select a "selection Variant" and at the bottom of the Report input screen, I choose an User- Specific "Layout Variant" which i have saved with few columns only to show in the report.
When i run this in foreground, i am able to get the report with my "Layout variant" with only my selected columns. I could also export this output and it looks fine in excel.
When i run the same report with the same "Layout Variant" in background and save it to my local desktop as an excel file, my "Layout Variant" doesn't seem to work and i am getting all the columns and fields of the reports as such.
Is this the way, the "Layout variants" will work? Whether this will not work in background jobs? Is this the way standard SAP work?
Please guide me...
Good rating points will be given for useful inputs..
Anand Viswanathan

Standard SAP functionality

Firefox does not start in Windows Vista. It can be seen in the task manager that the process is running. But nothing happens. After deleting the process it still does not work.

Are there any issues with Firefox 10 and Vista? IE works.

Hi webero,
There are no Known Issues that I am aware of. You should take a look at [[Firefox is already running but is not responding]] and our article called [https://support.mozilla.org/en-US/kb/Firefox-not-start Firefox will not start].
Hopefully this helps!

Can I install my old boot drive (Tiger) in another G4 running Leopard?

My trusty old dual 1GHz Quicksilver has finally died so I bought a used G4 Quicksilver MMD to replace it. Thing is, this other G4 has Leopard installed on it. To use Photoshop 7, I need to be running Tiger. If I install my old Tiger boot drive in this "new" box, can I choose to boot into either version of OS by simply holding down Option when starting up?
G4 1.25GHz Quicksilver Mirrored Drive Doors 1Gb RAM

Yes you can.
PhotoShop may need de-authorizing by phone to get it to work in the new Mac though... at least my CS3 was that way & still haven't done it.

I have completely taken off all of firefox from my application, and now when I try to load the firefox I receive a message that thereis another firefox running

before it crashed I was trying to take firefox to factory reset, my computer turned off before. I have a mac OSX 10.6

Check if you still have a profiles.ini file left if you've only deleted the Firefox profile folder:
* /Users/<user>/Library/Application Support/Firefox/
If you still have a profiles.ini file, but the profile folder is gone then delete the profiles.ini file to force Firefox to create a new default profile.
*https://support.mozilla.org/kb/Firefox+is+already+running+but+is+not+responding
*http://kb.mozillazine.org/Profile_in_use

Can OEM view jobs that has been scheduled by dbms_jobs

Hello,
Can I use OEM to view jobs that were scheduled using dbms_jobs. Thank you.

Hi,
If you mean the OEM client java console, then yes, you find it:
select db
then Distributed->Advanced Replication->Administration
you should then see a DBMS Jobs tab, which can be used to add/edit jobs.
cheers,
John

Identify Jobs that are not running on their Scheduled date time

I have 29 scheduled jobs that run at different intervals of time. Some run once a day. few others run on hourly basis while others run on Sundays.
I was working on a query that would let me know if a particular job did not run on its scheduled date and time.
SELECT * from all_scheduler_jobs WHERE state <>'DISABLED'; will give me a list of all jobs that I have to monitor and that are not in the disabled state. But how can I verify that the jobs are running at their scheduled date time?
Any help please? I need to create a view of all such jobs and then plan to send an alert so that appropriate action can be taken and it is assured that all important jobs run as per schedule.
Thanks.

Hi,
I can see 2 approaches.
- for jobs that have run but ran very late you should query dba_scheduler_job_run_details and filter by the difference between req_start_date and actual_start_date
- for jobs that should have run but shouldn't, query for DBA_SCHEDULER_JOBS jobs that are SCHEDULED where next_run_date is in the past
Hope this helps,
Ravi.

Manually running a cron job in SunOS 5.8

I was looking in the Solaris documentation and UNIX Man pages, however how do I manually run a cron job that is already set-up? I am troubleshooting an issue and the cron job runs just fine during the night, I would like to execute it manually for testing purposes. Thank you in advance.

Run this command :
crontab -l
And, look for the command you want to run and then just execute it on command line. The first five columns (separated by spaces) are date/time/day (skip those), after that you will see full path of the script/program. Just copy and paste that full path on command line.

Will Sql Server Agent start job when previous instance is running?

Just as an example, I have a job that is set to run every 5 minutes. The job takes 30 minutes to complete. So sql server agent executes the job at 9:30AM. Will the job run again at 9:35AM even though the previous instance hasn't completed?
In my case, I would like to reproduce what Task Scheduler already has, which is to
not begin a new instance until the previous instance has completed. When the previous instance completes, then 5 minutes later, the job will be run again.
Thanks.
VM

Apparently, it will skip it until the job is completed: http://stackoverflow.com/questions/865042/will-a-sql-server-job-skip-a-scheduled-run-if-it-is-already-running
VM

Account is SysAdmin - but Non-SysAdmins have been denied permission to run DTS Execution job steps without a proxy account.

Hi,
I have a SQL Agent Job that executes an SSIS package. However when I manually execute the SQL Agent job from SSMS I get the following seemingly common error:
"Non-SysAdmins have been denied permission to run DTS Execution job steps without a proxy account."
However, the job owner is a member of both the SQL SysAdmin role, and also a member of the operating system's Administrators group.
The SQL Agent Service is running under Local Service (not something within my control unfortunately).
I have found many posts suggesting setting up a proxy account etc, but that seems to be for the case where the job owner is not a sys admin. To be clear, the job owner IS a sysadmin.
This is occurring in the Production environment. I have two other, seemingly identical environments that are working fine, so i am wondering if I have missed something obvious.
Any idea what I've missed?
details:
Windows Server 2008 R2 Standard
SQL Server 2008 R2 (SP2) - 10.50.4000.0 (X64) Enterprise Edition
Note - the jobs were migrated to the broken environment by scripting them from Env 1, with If exists, drop, create etc.
Thanks
Guy
Guy

Obviously, as soon as I post this I figure out the issue:
In the script i had provided for migrating the job to the new environment (to which I have no access) the DBA replaced the @owner_login_name value with a different login. The account he used does have SysAdmin privileges but for some reason
changing the account in the script did not correctly update in the job.
On opening the job in SSMS it appeared to have the correct owner account, but using GUI to look up the specific account and adding it this way, while appearing to make no difference, has in fact solved the problem.
My assumption therefore is that updating the @owner_login_name when scripting the job is not the same as actually editing the job owner itself.
Guy
Guy

Run OS Command job return code

I have a 'Run OS Command' job that shows the correct output and performs the OS functions (trimming logfiles), but EM always says the job 'failed'.
I have tried forcing the script return code to 0 and non-0 values, but that does not make any difference.
How can I get EM to show a "success" status and what determines success/failed OS jobs?
James

Don,
There is a semantic difference when you stop a job or when you stop a specific execution of the job. User can stop a job from job activity page stop button or using "Stop Job" button on job results page. User can stop a execution by using "stop" button after selecting a job execution in specific job results page.
In former case job engine would stop the whole job and job will not be scheduled in future. However when you stop a specific execution of job, then job engine consider it as user is intending to stop only the current execution of the job so future run of job will continue to work as is.
I agree from end user perspective - the message on the confirmation screen is not very clear and we will file a bug and fix in next scheduled version of Enterprise manager.
-- Ashutosh

Scheduled SAP Job not executed as per restrictions

Hi All,
A job was scheduled to execute every Saturday (Job Frequency - weekly). Also in the Start time - Restrictions, the check box for "Execute only on Workdays" is checked.
Last saturday 27/08 was a holiday (displayed properly in the factory calendar), So the job should not have got executed.
But it got executed. I cannot find any reason for this.
Is there any other conditions that need to be provided so that it don't get executed on non-working days?

- Check in job definition if the factory calendar of the job is the good one (TBTCO-CALENDARID for the job that executed saturday), check also if job definition was changed (TBTCO-LASTCHDATE, LASTCHTIME and LASTCHNAME for next scheduled job) (or "Job details" on SM37)
- Check in the factory calendat, if the holiday was changed after saterday (SCAL, Extras, Display change docs.)
Regards,
Raymond

Can a dbms_job.run execute a job that execute another dbms_job.run?

Similar Messages

Maybe you are looking for