Operating system name and version: SLES10 SP3
Version of smsd: 3.1.14
Smsd installed from: sources
Name and model of a modem / phone: MTCBA-G-EN-F4-ED-EU (Modem with Accessory Kit) MultiModem GPRS Class 10, Quad Band (Ethernet)
Interface: TCP/IP

Have sometimes (on two servers) the same problem. It seems to be that the daemon after writing some errors will stop working with the modem.
See error log entries at bottom. At the top of these lines you see the first errors that called the alarm handler. I have a monitoring script that will restart the daemon in this case, so there was the first daemon restart right after the first lines.
After these messages (dont know how fast after) smsd stops talking to the modem but still move new files from outgoing to checked folder.

Doing a ps -ef|grep smsd |grep -v grep will show not only a line showing the running daemon but also a line containing:
[smsd] <defunct>
which means there was a broken task...

I have to restart sms daemon than to get it working again.

But I have no idea what maybe the cause...

error-log:
2011-01-27 13:36:36,7, GSMLAN1: Command is sent, waiting for the answer 2011-01-27 13:36:36,7, GSMLAN1: <- +CMS ERROR: 512 (MM establishment failure / weak signal, loosing connection) 2011-01-27 13:36:36,3, GSMLAN1: The modem answer was not OK: +CMS ERROR: 512 (MM establishment failure / weak signal, loosing connection) 2011-01-27 13:36:36,7, GSMLAN1: Running alarmhandler: /usr/ksp/script/root/smsalert.sh ALARM 2011-01-27 13:36:36 3 GSMLAN1 "The modem answer was not OK: +CMS ERROR: 512 (MM establishment failure / weak signal, loosing connection)" 2011-01-27 13:36:36,7, GSMLAN1: Done: alarmhandler, execution time 0 sec., status: 0 (0) 2011-01-27 13:36:36,3, GSMLAN1: Exec: alarmhandler said something: 2011-01-27 13:36:36,3, GSMLAN1: ! ALARM 2011-01-27 13:36:36 3 GSMLAN1 The modem answer was not OK: +CMS ERROR: 512 (MM establishment failure / weak signal, loosing connection) 2011-01-27 13:36:36,5, GSMLAN1: Waiting 10 sec. before retrying 2011-01-27 13:36:46,6, GSMLAN1: Checking if modem is ready 2011-01-27 13:36:48,5, GSMLAN1: Everything ok now. 2011-01-27 13:40:16,5, smsd: Outgoing file checker has started. PID: 16081. 2011-01-27 13:40:16,5, smsd: Everything ok now. 2011-01-27 15:02:30,7, GSMLAN1: -> AT 2011-01-27 15:02:30,7, GSMLAN1: Command is sent, waiting for the answer 2011-01-27 15:02:35,7, GSMLAN1: put_command expected (OK)|(ERROR), timeout occurred. 1. 2011-01-27 15:02:35,7, GSMLAN1: <- 2011-01-27 15:02:35,7, GSMLAN1: -> . 2011-01-27 15:02:35,7, GSMLAN1: Command is sent, waiting for the answer 2011-01-27 15:02:40,7, GSMLAN1: put_command expected (OK)|(ERROR), timeout occurred. 2. 2011-01-27 15:02:40,7, GSMLAN1: <- 2011-01-27 15:02:41,7, GSMLAN1: -> AT 2011-01-27 15:02:41,7, GSMLAN1: Command is sent, waiting for the answer 2011-01-27 15:02:46,7, GSMLAN1: put_command expected (OK)|(ERROR), timeout occurred. 3. 2011-01-27 15:02:46,7, GSMLAN1: <- 2011-01-27 15:02:46,7, GSMLAN1: -> . 2011-01-27 15:02:46,7, GSMLAN1: Command is sent, waiting for the answer 2011-01-27 15:02:51,7, GSMLAN1: put_command expected (OK)|(ERROR), timeout occurred. 4. 2011-01-27 15:02:51,7, GSMLAN1: <- 2011-01-27 15:02:51,7, GSMLAN1: -> AT 2011-01-27 15:02:51,7, GSMLAN1: Command is sent, waiting for the answer 2011-01-27 15:02:57,7, GSMLAN1: put_command expected (OK)|(ERROR), timeout occurred. 5. 2011-01-27 15:02:57,7, GSMLAN1: <- 2011-01-27 15:02:57,7, GSMLAN1: -> . 2011-01-27 15:02:57,7, GSMLAN1: Command is sent, waiting for the answer 2011-01-27 15:03:02,7, GSMLAN1: put_command expected (OK)|(ERROR), timeout occurred. 6. 2011-01-27 15:03:02,7, GSMLAN1: <- 2011-01-27 15:03:02,7, GSMLAN1: -> AT 2011-01-27 15:03:02,7, GSMLAN1: Command is sent, waiting for the answer 2011-01-27 15:03:07,7, GSMLAN1: put_command expected (OK)|(ERROR), timeout occurred. 7. 2011-01-27 15:03:07,7, GSMLAN1: <- 2011-01-27 15:03:08,7, GSMLAN1: -> . 2011-01-27 15:03:08,7, GSMLAN1: Command is sent, waiting for the answer 2011-01-27 15:03:13,7, GSMLAN1: put_command expected (OK)|(ERROR), timeout occurred. 8. 2011-01-27 15:03:13,7, GSMLAN1: <- 2011-01-27 15:03:13,7, GSMLAN1: -> AT 2011-01-27 15:03:13,7, GSMLAN1: Command is sent, waiting for the answer 2011-01-27 15:03:18,7, GSMLAN1: put_command expected (OK)|(ERROR), timeout occurred. 9. 2011-01-27 15:03:18,7, GSMLAN1: <- 2011-01-27 15:03:18,7, GSMLAN1: -> . 2011-01-27 15:03:18,7, GSMLAN1: Command is sent, waiting for the answer 2011-01-27 15:03:24,7, GSMLAN1: put_command expected (OK)|(ERROR), timeout occurred. 10. 2011-01-27 15:03:24,7, GSMLAN1: <- 2011-01-27 15:07:24,7, GSMLAN1: -> AT 2011-01-27 15:07:24,7, GSMLAN1: Command is sent, waiting for the answer 2011-01-27 15:07:29,7, GSMLAN1: put_command expected (OK)|(ERROR), timeout occurred. 1. 2011-01-27 15:07:29,7, GSMLAN1: <- 2011-01-27 15:07:29,7, GSMLAN1: -> . 2011-01-27 15:07:29,7, GSMLAN1: Command is sent, waiting for the answer 2011-01-27 15:07:35,7, GSMLAN1: put_command expected (OK)|(ERROR), timeout occurred. 2. 2011-01-27 15:07:35,7, GSMLAN1: <-

Axel

CMS 512 is a problem with GSM network, not a problem with communication between modem and smsd. If your script does not restart smsd after this error, how does the communication continue?

In my case, when a modem stops responding, I track the issue with the following alarmhandler:



After modem has stopped responding, restarting of smsd is not enough. The system needs to switch power off-on to get the modem working again.

In your case after smsd was restarted, the communication worked more than one hour. From the log you could check how the modem process has stopped. I assume that when smsd tried to close and re-open the port, opening has failed (with some error code and explanation), and because of this, the modem process has stopped. In this case the alarmhandler is also called, and you should track that message and try to restart the daemon.

[smsd] <defunct> is shown in the process list when modem process has stopped, but mainprocess still continues and spools messages. This is very abnormal case, and some external watchdog can be used to detect this kind of a zombie from the process list.

What kind is your current configuration? And do you run two servers simultaneously?