C-119
Cisco Media Gateway Manager 5.0 User Guide
OL-5461-02
Appendix C Troubleshooting
Miscellaneous Problems
b. If the DMD received the message but it is not reflected in the database. Identify which databroker
module dropped the ball. First check the DMD to see if it forwarded the message. If it didn't check
for the reason. The format may be wrong etc. one easy search/grep is node/slot/port/vpi/vci. This is
output every time a cache query is done.
grep "DROP MSG: validation error" dmd<dmd_id>.<pid>.* - prints out dropped messages.
c. If the message made it to DMD, we need to see if the sdbroker received the message.
grep "node <node_id> slot <slot#> .* port <logical port> sCh1 <vpi/dlci> sCh2 <vci>"
sdbroker*Msg*
d. If it looks like the message was received by the sdbroker but the database does not reflect the change,
then we should verify if there is an inconsistency between the databroker Cache and the database.
To dump the cache enter:
$ psg sdbroker - the process ID of the sdbroker will be returned
$ kill -USR1 <process id returned from the previous command>
The cache dump will be written to a file in /opt/svplus/log/sdbkrCache.dump Verify if the connection in
question is correct in the cache.
Defect Information—We need the logs of the processes on both sides of the interface which the message
was dropped. A dump of the specific user_connection table entry that is incorrect is also useful as well
as the segment tables for this connection. The specific node, slot, logical_port, vpi, vci of both ends of
the connection in question. If a cache dump was done, include that also.
Possible alternative workarounds—If the problem is between the sdbroker cache and the database, a
cache resync can be done. TO do this execute the command /usr/usrs/svplus/tools/CacheResync [-t
<time in seconds>]. If the problem still exists, collect the defect information.
Related key index entries: inconsistency, connection, databroker
C.15.2.2 Inconsistent Connection Status
The GUI display of 'Status' is from the 'state' field of the user_connection table. The field represents the
worst condition of any of the segments in the connection. This field values are:
1 = clear
2 = fail
3 = Down
4 = incomplete
5 = Error
Step 1 If the value is not what is expected, check the connection level databases for each segment to see if they
are correct.The last message on the segment in question is the important one.
Step 2 If the status is 'incomplete" it means that one of the segments of the connection is not in the
user_connection table, check the "in_seg" flags in the user_connection table entry for this connection.
Search the database and messages for the segment that has flag = 0. Also see next step on Error
connection.
Step 3 If two master endpoints each have the same slave endpoint, then we have an errored connection. The
first connection established will have a status of "error' and the second master endpoint connected to the
slave will have a status of 'incomplete".
Comments to this Manuals