| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 2894.1 | more information? | WHOS01::ELKIND | Steve Elkind, Digital SI @WHO | Tue Jun 03 1997 16:31 | 2 |
| What version is the "old" version? Did they rebuild the application
running on the node with the v3.2A-22 version?
|
| 2894.2 | Precisions | EVTSG8::$VOGEL | | Wed Jun 04 1997 11:17 | 23 |
| hello,
The old version of DMQ on the node was v 3.2 (25).
The extraction program is not running on the node where DMQ version has
changed. In fact, generating a new executable for extraction has been
done in life environnement.
May be I was not clear with my initial explanations. On one side we
have a cluster (cluster A with 2 nodes (1 node with version v 3.2 (25)
of DMQ and on the other node v 3.2A-22) and on the other hand another cluster
(cluster B). Extraction prog is running on one of the nodes from
cluster B and is accessing a RDB database on cluster A. When we point
to node 1 of cluster A with old version of DMQ it works,
when we point to node of cluster A
with new DMQ version extraction prog blocks (does not go on, no error
message, no cpu, no IO ).
DO you have any idea what we could do ?
Thank you for your help.
Kind regards. Georges.
|
| 2894.3 | Find out what you can do... | PAMSRC::MICHELSEN | BEA/DEC MessageQ Engineering | Wed Jun 04 1997 12:48 | 14 |
| re: .2
...by trying different MQ operations (i.e. looping messages, monitoring message counts...). Also, you need to
look at the EVL log file to see if the messages are being lost. I would then look at the problem process with SDA
to see if it's hung in an unusual process state (i.e. MWAIT), MQ programs wait for messages in LEF state. Also,
in SDA, look and see if ASTs are turned off or the process is stuck in an AST. From the DMQ$MGR_UTILITY you can
enable tracing on your process (assuming USER mode ASTs can get through) and watch the messaging traffic to the
process.
Hopefully, the results of the info gathered will lead you to narrowing your problem.
Marty
|
| 2894.4 | for those with tunnel (80-col) vision | WHOS01::ELKIND | Steve Elkind, Digital SI @WHO | Wed Jun 04 1997 17:38 | 24 |
| <<< PAMSRC::SYS$SYSDEVICE:[NOTES$LIBRARY]DECMESSAGEQ.NOTE;1 >>>
-< NAS Message Queuing Bus >-
================================================================================
Note 2894.3 Prob.access.remote RDB db due to DMQ3.2A-22 3 of 3
PAMSRC::MICHELSEN "BEA/DEC MessageQ Engineering" 14 lines 4-JUN-1997 11:48
-< Find out what you can do... >-
--------------------------------------------------------------------------------
re: .2
...by trying different MQ operations (i.e. looping messages, monitoring message
counts...). Also, you need to look at the EVL log file to see if the messages
are being lost. I would then look at the problem process with SDA to see if
it's hung in an unusual process state (i.e. MWAIT), MQ programs wait for
messages in LEF state. Also, in SDA, look and see if ASTs are turned off or
the process is stuck in an AST. From the DMQ$MGR_UTILITY you can enable
tracing on your process (assuming USER mode ASTs can get through) and watch the
messaging traffic to the process.
Hopefully, the results of the info gathered will lead you to narrowing your
problem.
Marty
|