Conference azur::mcc

3149.0. "various EXPORTER error messages" by ANNECY::BONNIER (Jean-Luc ... EIC/T&N Annecy France) Tue Jun 09 1992 11:20

This topic is related to 3110 which got no reply yet.

It gives some precisions about MCC_EXPORTER_BACKGROUND.LOG messages which
I don't understand.
I remind you that we are running several exporter jobs at a time and possibly
on the same entities.
Each of these exporter jobs creates its own Rdb file.

For example I had two exporter jobs running in parallel and the machine they
were running on (VMS 5.5, BMS T1.2.7) was shutting down and rebooted.

One job is aborted due to:
$       BTS == "$SYS$SYSTEM:MCC_EXPORTER_FM_BG.EXE"
$       BTS "NSDP$ROOT:[NSDP_REPORT.PER1_W2392.WORK]NSDPRN_EXPORT.RDB" ""
%Internal error: enter_kernel: deadlock
%CMA-F-BUGCHECK, internal error detected in DECthreads

Another job is aborted due to:
$       BTS == "$SYS$SYSTEM:MCC_EXPORTER_FM_BG.EXE"
$       BTS "NSDP$ROOT:[NSDP_REPORT.LNG_W2392.WORK]NSDPRN_EXPORT.RDB" ""
%JBC-F-QUERESET, job aborted during execution because queue was reset

Why these two messages are different ?
What means the first one ?

-------------------------------------------------------------------------------

With the same type of configuration,
another strange exporter message has been collected.

$ MANAGE/ENTERPRISE
DECmcc (T1.2.7)

EXPORT NODE4 STKRT6 CIRCUIT GOTEBORG  END TIME = 24-MAY-1992 23:59:59 ,EXPORT
PERIOD = 01:00:00.0 ,EXPORT TARGET = NSDP$ROOT:[NSDP_REPORT.UGO_W2192.WORK]NSDPRN_EXPORT.RDB

Node4 STKRT6 Circuit GOTEBORG
AT 18-MAY-1992 00:04:00


The requested operation cannot be completed
	    MCC Routine Error = %MCC-E-INVREPOS, invalid Repository ID,
	Entity Existence Info = Entity Existence Cannot Be Determined
     
What does it mean ?
We suspect that the EXPORT command is issued before the background process
is properly initialzed. Is it possible ? 
If YES, is it enough to check the presence of the 
batch job in Executing state before doing EXPORT commands ?
OR
If YES and not enough, what is the time to wait before doing EXPORT commands ?


Thanks for your help.
Jean-Luc

Our exporter experts are out of the office until next week, so it may
be a bit longer before you see an answer.

Are the exporter experts back ?

I am strongly waiting for your support.

Thanks again.
Jean-Luc

    

> "%Internal error: enter_kernel: deadlock
>%CMA-F-BUGCHECK, internal error detected in DECthreads"

It tells that a dead lock situation is detected. In T1.2.7 there is a bug
in Exporter's locking mechanism. Usually it happens when the user specifies
Delete Export or Modify Export directive in the time when the background 
process is its restart procedure. It's fixed in 1.2.21.

>%JBC-F-QUERESET, job aborted during execution because queue was reset

I do not know what this error is about. Something with the queue where you
run the background process?

>The requested operation cannot be completed
>	    MCC Routine Error = %MCC-E-INVREPOS, invalid Repository ID,
>	Entity Existence Info = Entity Existence Cannot Be Determined
>     
>What does it mean ?
>We suspect that the EXPORT command is issued before the background process
>is properly initialzed. Is it possible ? 

	No. It's not related to the background .
Usually it tells that your MIR is corrupted. Again one the MIR problem which
can give this behavior is fixed after T1.2.7.

>... what is the time to wait before doing EXPORT commands ?

	Use the latest version of the MCC and you do not need to wait at all.

	In T1.2.7 there were several bugs which can give "Backgorund 
exporting process is not set up" error.
If the background process is alive it's enough to repeat the command.
All this problems are fixed in 1.2.21.

	Sam Shmuylovich

    

Thank you Sam.

I didn't know there was a MCC 1.2.21
On my side, I'll install the SSB version since it is released soon.

Jean-Luc