| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 7188.1 | Is this still reproducable ? | HANNAH::KAMATH | | Fri May 02 1997 12:24 | 13 |
| Hi,
Is this problem reproducable, have you observed any pattern
like occurances of error/warning messages before crashing.
How many active queues exist on the system ?
What was the queue manager image prior to installation of the
patch ?
Does this occur with only LTA type of devices ?
-Harish Kamath.
|
| 7188.2 | Running V1.3 and a different printer triggers it | NECSC::HARVEY | Printserver Support- America's Zone | Fri May 02 1997 12:42 | 37 |
| harish,
The customer is now running DCPS V1.3 version and another or
different printer seems at fault this time. Notice:
%%%%%%%%%%% OPCOM 2-MAY-1997 08:03:52.01 %%%%%%%%%%%
Message from user SYSTEM on DNC
Queue DNC_NSPLASER2: %DCPS-I-WARMUP, Print Engine is in warmup state
%%%%%%%%%%% OPCOM 2-MAY-1997 08:04:03.35 %%%%%%%%%%%
Message from user SYSTEM on DNC
Queue DNC_NSPLASER2: %DCPS-I-WARMUP, Print Engine is in warmup state
%%%%%%%%%%% OPCOM 2-MAY-1997 08:35:17.90 %%%%%%%%%%%
Message from user SYSTEM on DNC
Queue DNC_NSPLASER2: %DCPS-I-WARMUP, Print Engine is in warmup state
%%%%%%%%%%% OPCOM 2-MAY-1997 08:52:36.58 %%%%%%%%%%%
Message from user SYSTEM on DNC
Process SYMBIONT_32: %SYSTEM-F-ACCVIO, access violation, reason
mask=00,
virtual address=00000000, PC=000DF5D1, PSL=0BC00000
%%%%%%%%%%% OPCOM 2-MAY-1997 08:52:37.04 %%%%%%%%%%%
Message from user QUEUE_MANAGE on DNC
%QMAN-E-SYMDEL, unexpected symbiont process termination
%%%%%%%%%%% OPCOM 2-MAY-1997 08:52:37.05 %%%%%%%%%%%
Message from user QUEUE_MANAGE on DNC
-SYSTEM-F-ACCVIO, access violation, reason mask=00, virtual
address=00000000,
PC=0001D764, PSL=03C00000
So it seems that it is only LTA devices that trigger the symbiont
crashes.
Renis
|
| 7188.3 | Do you find the DCPS and queue manager dumps. | HANNAH::KAMATH | | Fri May 02 1997 15:58 | 31 |
| Renis,
You seem to be getting a different error now and still access
violating. Can you get the following information, this is
essential to understand the cause of this problem.
1. What is your hardware ?
2. What was the previous queue manager ident ?
3. Is this is a cluster member ?
- If so how is queue manager configured ?
- Is QM running on the same node as queues are running.
4. How many active DCPS queues exist on the system ?
5. Check if there exist DCPS symbiont,queue manager or
job controller in sys$specific and in sys$common areas.
- If so delete the old files and retest.
6. Check for DCPS and Queue Manager and/or Job controller dump
files in sys$common and sys$specific area.
7. Collect the analyze output of DCPS,QUEUE Manager and/or Job
controller
ANALYZE/PROCESS/IMAGE=<image name> <corresponding
dump filename>/full
> SHOW IMAGE
> SHOW CALLS
> SHOW STACK
I need answers to all the above questions to help you further.
-Harish Kamath.
|
| 7188.4 | More answers | NECSC::HARVEY | Printserver Support- America's Zone | Fri May 02 1997 21:32 | 99 |
| Harish,
Here are the answers to your questions:
> 1. What is the hardware ?
It's a VAX 3100-95, VMS 6.1
> 2. What was the previous queue manager ident ?
I don't understand "previous" here. It was upgraded from VMS 5.5-2
long, long ago. The current QUEMAN ident is below. The patch I
mentioned was installed long ago, not recently.
Image Identification Information
image name: "QUEMAN"
image file identification: "X-8"
link date/time: 9-MAR-1994 02:04:03.71
linker identification: "05-13"
> 3. Is this is a cluster member ?
> - If so how is queue manager configured ?
> - Is QM running on the same node as queues are
running.
No, it is a standalone system.
> 4. How many active DCPS queues exist on the system ?
Nine.
> 5. Check if there exist old versions of the DCPS symbiont,
> VMS queue manager or job controller in sys$specific and in
> sys$common areas.
> - If so delete the old files and retest.
Good thought, I hadn't looked, but nope, none.
> 6. Check for DCPS and Queue Manager and/or Job controller dump
> files in sys$common and sys$specific area.
> - Provide copies of them if possible.
I have one:
Directory SYS$SYSDEVICE:[SYS0.SYSEXE]
DCPS$SMB.DMP;1 13551/13551 29-APR-1997 11:00:05.91 [SYSTEM]
This is before the V1.2 -> V1.3 upgrade. I can copy it up if you
give me whatever authorization I need to send a big file, or I can
FTP it somewhere if you like.
> 7. Collect and then analyze output of DCPS,QUEUE Manager and/or
Job
> controller.
> ANALYZE/PROCESS/IMAGE=<image name> <corresponding
> dump
filename>/full
> > SHOW IMAGE
> > SHOW CALLS
> > SHOW STACK
From the above dump file, I cannot. Here is what happens (do note that
the old DCPS$EXE is not present on the system, if that matters, since
the dump was from V1.2). Variations of /image and not have no impact.
$ ana/proc/image dcps$smb.dmp
Creating a subprocess
Condition signalled to take dump:
%SYSTEM-F-ACCVIO, access violation, reason mask=01, virtual
address=20000000,
PC=80000010, PSL=03C00004
%SYSTEM-F-ACCVIO, access violation, reason mask=01, virtual
address=20000000,
PC=80000010, PSL=03C00004
%DEBUG-I-CANTCREATEMAIN, could not create the debugger subprocess
%DEBUG-I-CANTCREATEMAIN, could not create the debugger subprocess
-LIB-F-NOCLI, no CLI present to perform function
-LIB-F-NOCLI, no CLI present to perform function
%DEBUG-I-SHRPRC, debugger will share user process
%DEBUG-I-SHRPRC, debugger will share user process
Let me know if you need anything else, or want this dump file. NOte
that it is not producing a dump file for each unexpected termination.
There have been no more unexpected symbiont terminations since changing
to single threaded operation.
Thanks renis
|
| 7188.5 | Copy the dump file to HANNAH:: | HANNAH::KAMATH | | Mon May 05 1997 11:53 | 14 |
| Renis,
Copy the dump file to HANNAH:: , will try to analyze with DCPS V1.2
images on our system.
Do you find any other dumps created on the system ? If so, where
these created when DCPS dump was generated.
The QUEMAN ident given in .4 update is not of any help, can you
get the image ident of QMAN$QUEUE_MANAGER.EXE file.
DCPS should work normally in multi-stream configuration.
-Harish Kamath.
|
| 7188.6 | Dump is corrupt and the QMAN is: | NECSC::HARVEY | Printserver Support- America's Zone | Mon May 05 1997 16:00 | 16 |
| Harish,
The dump file is corrupt. The customer can not copy it via DSN link
or FTP it to us. Here is the qman data:
Image Identification Information
image name: "QMAN$QUEUE_MANAGER"
image file identification: "X-3"
link date/time: 16-AUG-1996 19:58:48.58
linker identification: "05-13"
Renis
|
| 7188.7 | He is running single stream now | NECSC::HARVEY | Printserver Support- America's Zone | Mon May 05 1997 16:02 | 6 |
| Harish,
I forgot to add that the customer has not seen any aborts/crashes
since going to single stream mode.
Renis
|
| 7188.8 | Problem not reproducable. | HANNAH::KAMATH | | Wed May 07 1997 12:57 | 19 |
| Renis,
I tried reproducing the posted problem by installing
the queue manager patch on OpenVMS VAX V7.0 system
and with DCPS V1.2. DCPS stream value was set to 4.
The printer used was DEClaser 2250.
I was not able to reproduce the condition on our
system.
Do you know what kind of print jobs can force this
condition ? Is this seen only when printer reports
an information/error message to the host ? Do you
know if a specific job or queue operation that will
force this condition ?
-Harish Kamath.
|
| 7188.9 | Special tools/messages to capyure better data? | NECSC::HARVEY | Printserver Support- America's Zone | Thu May 15 1997 15:32 | 13 |
| Harish,
Well the custoemr has been running in single stream mode for a
while and have not suffered any aborts or symbiont crashes. His users
don't want him to go back to multistreamed mode however he needs to
balance the use of VMS resources, process slots, etc.
Is there any special debugging messages, features that can be enabled
to capture more information that just the normal opcom messages? The
custoemr needs to go back to multistream symbionts.
Renis
|
| 7188.10 | Do you have a dump file ? | HANNAH::KAMATH | | Fri May 16 1997 11:04 | 12 |
| Renis,
Can you copy the dump file(s) to HANNAH:: .
Do you notice any queue manager and Job controller
error messages in operator log ? If so, can you post
these messages along with DCPS opcom messages with a
window of 5 as a reply to this note.
-Harish Kamath.
|