Dear Users
I have a strange issue that I can't seem to find the direct cause.
I have running SMPP Box, with many ESME connected fine. It seems that
when my Kannel Service is restarted many different ESME cannot rebind.
SMPP Box log displays the following for those attempting to connection:
2016-07-21 11:20:00 [9301] [2504] DEBUG: type_name: bind_transceiver
2016-07-21 11:20:00 [9301] [2504] DEBUG: command_id: 9 = 0x00000009
2016-07-21 11:20:00 [9301] [2504] DEBUG: command_status: 0 =
0x00000000
2016-07-21 11:20:00 [9301] [2504] DEBUG: sequence_number: 1329 =
0x00000531
2016-07-21 11:20:00 [9301] [2504] DEBUG: system_id: "*****"
2016-07-21 11:20:00 [9301] [2504] DEBUG: password: "*****"
2016-07-21 11:20:00 [9301] [2504] DEBUG: system_type: NULL
2016-07-21 11:20:00 [9301] [2504] DEBUG: interface_version: 52 =
0x00000034
2016-07-21 11:20:00 [9301] [2504] DEBUG: addr_ton: 1 = 0x00000001
2016-07-21 11:20:00 [9301] [2504] DEBUG: addr_npi: 1 = 0x00000001
2016-07-21 11:20:00 [9301] [2504] DEBUG: address_range: NULL
2016-07-21 11:20:00 [9301] [2504] DEBUG: SMPP PDU dump ends.
2016-07-21 11:20:00 [9301] [2504] DEBUG: SMPP[*****]: Sending PDU:
2016-07-21 11:20:00 [9301] [2504] DEBUG: SMPP PDU 0x7f70f001f950 dump:
2016-07-21 11:20:00 [9301] [2504] DEBUG: type_name:
bind_transceiver_resp
2016-07-21 11:20:00 [9301] [2504] DEBUG: command_id: 2147483657 =
0x80000009
2016-07-21 11:20:00 [9301] [2504] DEBUG: command_status: 0 =
0x00000000
2016-07-21 11:20:00 [9301] [2504] DEBUG: sequence_number: 1329 =
0x00000531
2016-07-21 11:20:00 [9301] [2504] DEBUG: system_id: "OurSytemId"
2016-07-21 11:20:00 [9301] [2504] DEBUG: SMPP PDU dump ends.
2016-07-21 11:20:00 [9301] [2504] ERROR: Invalid SMPP PDU received.
Many other clients can bind fine however a minority cannot. The bind
appears in the status page and then disconnects immediately.
When I run packet traces I can see the bind_transceiver_resp is sent
fine and is acknowledged.
I'm thinking there has to be something ins SMPPLogins file that is
different for these clients however nothing is obvious. All follow the
same format as clients connected ok.
The even more strange thing is that at some moment in time, 1-2 hours
after the service restart, all of those failing connections magically
bind ok. As if a timer or cache is preventing them all from
re-connecting for a certain period...
I've checked around the code source and previous issue but cannot find
anything obvious. I'd appreciate any ideas on this.
Thanks
Lee
