Operating system name and version: WIndows with cygwin
Version of smsd: 3.1.12
Smsd installed from: sources
Name and model of a modem / phone: Huawei E176
Interface: USB


Dear Keke:

Thanks for your post. I´ve updated the smsd to 3.1.12, and work fine with Cigwin, but I have some problems.

When I start SMSD, the modem works fine and send a lot of messages at a rate of 10 sms per second., With a message like this:

GSM1: -> 0011000C914511564746020000FF9FC3B79B5E9ED3C3A0F29C5E064D9B53D0F8ED0659AF2 C5 01E549C4283D320885E06BDC5F372BC9E0E8340E3B71B547783C8E5F9B85E76D3DF207219949BC1602001A8EC06D1EBA0E39309A24A8B4E22E80D9A56A541A70B04AABA41E17979FE9683C6EF76593E4E87D9A07919347EBBE9E1313D2C0F83DAF53C082E7FBBE96F501844AF83C665769D1D96BB00. GSM1: Command is sent, waiting for the answer GSM1: <- ^RSSI:3 +CREG: 1 +CGREG: 1 +CREG: 1 +CGREG: 1 ^RSSI:7 +CMGS: 234 OK GSM1: SMS sent, Message_id: 234, To: 541165746420, sending time 14 sec. GSM1: Deleted file /var/spool/sms/checked/541165746420.txt GSM1: I have to send 1 short message for /var/spool/sms/checked/541165746617.txt GSM1: Sending SMS from to 541165746617 GSM1: Checking if Modem is registered to the network GSM1: -> AT+CREG? GSM1: Command is sent, waiting for the answer GSM1: <- +CREG: 1,1 OK GSM1: Modem is registered to the network GSM1: -> AT+CMGS=154

After 15 or 20 minutes, show the following message and after that, the rate became slower ( two to four messages per minute)

I stop the application, reset the modem, and restart the application and repeat the cycle.

¿How can I fix that?

After that, with your comments, I have a question. ¿you think that If I change to Linux, I have a very better performance instead cygwin?

¿or I have another possibilities to improve the performance with cygwin?

Thanks for your answers

Lextor



-> 0011000C914511156124840000FF9DC3B79B5E9ED3C3A0F29C5E064D9B53D0F8ED0689AC5 Command is sent, waiting for the answer <- ^MODE:3,3 ^MODE:3,3 ^MODE:3,3 +CMS ERROR: 500 (Unknown error) The modem answer was not OK: ^MODE:3,3^MODE:3,3^MODE:3,3+CMS ERROR: 500 (Unk Waiting 10 sec. before retrying Checking if modem is ready Unexpected input: ^MODE:3,3

lextor wrote

I´ve updated the smsd to 3.1.12, and work fine with Cigwin, but I have some problems.

Just for curiosity, you did not have any problems with the previous version of smsd? (which was 3.1.7, or?)

lextor wrote

When I start SMSD, the modem works fine and send a lot of messages at a rate of 10 sms per second.

How many modems do you currently have? Or did you mean "10 sms per minute" or "1 sms per 10 seconds"?

lextor wrote

After 15 or 20 minutes, show the following message and after that, the rate became slower ( two to four messages per minute)

I stop the application, reset the modem, and restart the application and repeat the cycle.

¿How can I fix that?

When the sending fails, is the message always something like:
<- ^MODE:3,3 ^MODE:3,3 ^MODE:3,3 +CMS ERROR: 500 (Unknown error)
...including some MODE?


I'm not sure, but perhaps your modem does not work well with the speed of communication which was changed in the version 3.1.9. To check this, apply the following settings for a modem(s):

send_delay = 1
send_handshake_select = no

lextor wrote

After that, with your comments, I have a question. ¿you think that If I change to Linux, I have a very better performance instead cygwin?

¿or I have another possibilities to improve the performance with cygwin?

For actual sending and receiving there is no remarkable difference.

The problem in Windows / Cygwin is a file system which is very slow when there are lot of files in the single directory. Linux works very well with 10000 - 15000 files in the checked directory, but in Cygwin some 500 - 1000 files will cause remarkable delays when smsd is searching for a file to send. Your front-end application should ensure that not too many files are placed into the spooler at the same time.

Current version 3.1.12 will show in the log file if a file system is going too slow. If getting a file takes more than one second, it is logged with a level DEBUG. And if it takes 5 seconds or more, the level used is NOTICE. The message contains also number of SMS files and number of .LOCK files seen, which helps when determining how many messages the file system can handle.

Dear Keke

There is my answers

1 - My problem is always the same. I can´t send SMS with various modems without a loss of speed (1 modem= 6 or 8 sms per minute, 3 modems = 1 or 2 SMS per minute per modem). For that, I don´t have advantage to aggregate any modem.

2 - I have two computers. One with 1 modem and other with 2 or 3 modems
Now, both with the last version of SMSD. The log file I send you, correspond to the computer with only one modem.

3- I obtain 1 sms per 6 or 10 seconds (that is fine for me)

4 - Yes, I have "Mode"s. You can see that in the log file
5 - Ok . I will do that , and tell you the results.

Before your answer, I read in the community, different issues about "ERROR 500". I found the possibility of inclusion of "memory_start =0"

I didnt know if it applies to my problem, but I changed that, and seems work much better.

Now, I include your statements and tell you the results (in both machines)

Thanks for your help
Lextor

lextor wrote

1 - My problem is always the same. I can´t send SMS with various modems without a loss of speed (1 modem= 6 or 8 sms per minute, 3 modems = 1 or 2 SMS per minute per modem). For that, I don´t have advantage to aggregate any modem.

If with 3.1.12 and memory_start = 0 you still have this loss of speed, you could send a large log with loglevel = 7 to me, so I can examine it.

Instructions in a Private block:

Hidden private text.

lextor wrote

4 - Yes, I have "Mode"s. You can see that in the log file

My Huawei E1552 does not show those MODE's. It shows some other messages (boot reports), and I can get rid of them with init2 = AT^BOOT=0,0 setting. You could try this too.

Could it be possible that your device is trying to connect to internet, and this causes the error?

lextor wrote

Before your answer, I read in the community, different issues about "ERROR 500". I found the possibility of inclusion of "memory_start =0"

I didnt know if it applies to my problem, but I changed that, and seems work much better.

If there is one message received, and lot's of memory spaces, it will take a while until all slots from 1...n are tried to read (when memory_start is not zero). But this should not cause error when sms is tried to send.

Dear Keke:

I send you (uploaded), a log file and a smsd conf file

As you can see, the errors decrease a lot, but the speed of sending , seems slowly

I wait for your comments

Thanks a lot

Lextor

Thanks, I will check those files deeply tomorrow morning. With a quick view you really have some troubles with your current setup...

By the way, the log is somehow modified, and has some tabs like spreadsheets do:

06/09/2010 09:31:52,5, GSM1: SMS sent, Message_id: 106, To: 5411XXXXXXXX, sending time 3 1 sec.

This prevents me from using some scripts which will help while analyzing the data. Also, the search does not work properly with your log. Can you provide an unaltered log for me?

Dear Keke:

I send you a completed (updated) and unaltered log file.

Thanks

Lextor

First, there are some delays with the filesystem, but those are not usual and not remarkable. It seems that Windows was just doing something, probably with a pagefile, and it caused a delay of few seconds. This is not the reason for your problem.

With one modem only everything worked fine, but not much messages were sent. I assume that with one modem sending much more messages also works fine. Then, when you send with two modems, both modems will fail sometimes, but usually retrying helps. This however causes the performance loss. It sounds like this is a problem in GSM network, and I cannot know how to fix it.

For some reason you have disabled check_network. It may speed up sending little bit, but you could enable it to see if there is a problem with registration. I know some modems which "drops" from the network when a message is received. When that kind of a modem is "dropped", sending will not work. Therefore smsd must wait until modem is registered again, and after that the sending works.

In your case the problem of course is, that much status reports are received continuously, and this may cause lot of collisions. And as I can see from the log, you really need those reports and disabling of them will not be an acceptable solution.

Sorry to say, but right now I'm unable to give any complete solution for you. It would be interesting to see how registration works when checking for it is enabled and two or more modems are sending simultaneously. If you run this, you could send a log to me.

After running with check_network enabled, you could also check if receive_before_send = yes causes any positive effect. Probably it does not, but because sooner or later status reports should be read from the modem, it may be good to read them as fast as possible.

Dear KekeÑ

Today I will make the changes and send a log for you.

But I don´t think this is the problem.

When I send SMS with two modems (one on a computer, one in another), the two modems works fine (at 6 or 8 SMS per minute each). The modems are in the same GSM network and receive the same amount of status reports

After that , I put the two modems in the same machine, and the speed decrease at 5 or 6 SMS in total (the two modems). Again, the network is the same and the amount of status reports are the same.

By the way, the two machines are P IV with 4 Gbs of ram.

With 3 modems , the amount of SMS sended , don´t increase ( never pass of 10 SMS per minute IN TOTAL)

For that, I´m working now with two modems (in different computers), because I haven´t any advantage to include anothe one.

This is not a normal operation for me.

Thanks a lot

Lextor

Ok, thanks for the clarification.

Make this change to the smsd.conf and try again with two modems in the same machine:

[default]
pre_init = no
init = ATE1+CMEE=1

After this change, everything is echoed back from the modem. When a PDU is sent, and it gets CMS 500, it is possible to check if the PDU is broken or not.

Dear Keke

I send you a log of today (with the modifications suggested)

with 3 modems

I wait for your comments.

Thanks

It seems that there are no any problems with the communication between PC and modems.

Have you already tried a setting init2 = AT^BOOT=0,0 or init2 = AT^BOOT=0 ? If not, you could try.

Do those sticks have an external antennas?

When you are running with three sticks on the same computer, how close they are to each other?

Are they connected directly to the computer, or are you using a hub? If using a hub, is it powered by an external power adapter which provides current enough?

Keke:

Ok, I try these commands now,

- The sticks hasn´t external antennas, they seems like memory sticks

- The computer is a laptop, there are two USB connectors in one side (with separation about 4 or 5 ", and the other to the other side of the laptop

- There are connected directly to the laptop, and there are no alarms of power . ( in other machines I receive an alarm that the power is not enough for that kind of element)

Lextor

Dear Keke:

¿ Have you another suggestion for me?

The message "ERROR 500" comes frecuently and the speed with two modems is very slow.

Thanks
Lextor

All sent PDU's are verified, and there was no any errors. You do not have to use the "echo" trick anymore.

When sending with single modem only, the timing is:

2010-09-06 09:09:08,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 09:09:10,7, GSM1: Took 2 sec. to get the answer for PDU. 2010-09-06 09:09:16,7, GSM1: Took 4 sec. to send a PDU. 2010-09-06 09:09:18,7, GSM1: Took 2 sec. to get the answer for PDU. 2010-09-06 09:09:24,7, GSM1: Took 4 sec. to send a PDU. 2010-09-06 09:09:26,7, GSM1: Took 2 sec. to get the answer for PDU. 2010-09-06 09:09:32,7, GSM1: Took 4 sec. to send a PDU. 2010-09-06 09:09:34,7, GSM1: Took 2 sec. to get the answer for PDU. 2010-09-06 09:09:40,7, GSM1: Took 4 sec. to send a PDU. 2010-09-06 09:09:42,7, GSM1: Took 2 sec. to get the answer for PDU. 2010-09-06 09:09:48,7, GSM1: Took 4 sec. to send a PDU. 2010-09-06 09:09:50,7, GSM1: Took 2 sec. to get the answer for PDU. 2010-09-06 09:09:56,7, GSM1: Took 4 sec. to send a PDU. 2010-09-06 09:09:58,7, GSM1: Took 2 sec. to get the answer for PDU. 2010-09-06 09:10:04,7, GSM1: Took 4 sec. to send a PDU. 2010-09-06 09:10:06,7, GSM1: Took 2 sec. to get the answer for PDU. 2010-09-06 09:10:12,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 09:10:14,7, GSM1: Took 2 sec. to get the answer for PDU. 2010-09-06 09:10:20,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 09:10:27,7, GSM1: Took 7 sec. to get the answer for PDU. 2010-09-06 09:10:34,7, GSM1: Took 4 sec. to send a PDU. 2010-09-06 09:10:41,7, GSM1: Took 7 sec. to get the answer for PDU. 2010-09-06 09:10:48,7, GSM1: Took 4 sec. to send a PDU. 2010-09-06 09:10:49,7, GSM1: Took 1 sec. to get the answer for PDU. 2010-09-06 09:10:55,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 09:10:57,7, GSM1: Took 2 sec. to get the answer for PDU. 2010-09-06 09:11:03,7, GSM1: Took 4 sec. to send a PDU. 2010-09-06 09:11:05,7, GSM1: Took 2 sec. to get the answer for PDU. 2010-09-06 09:11:11,7, GSM1: Took 4 sec. to send a PDU. 2010-09-06 09:11:13,7, GSM1: Took 2 sec. to get the answer for PDU. 2010-09-06 09:11:19,7, GSM1: Took 4 sec. to send a PDU. 2010-09-06 09:11:26,7, GSM1: Took 7 sec. to get the answer for PDU. 2010-09-06 09:11:34,7, GSM1: Took 4 sec. to send a PDU. 2010-09-06 09:11:36,7, GSM1: Took 2 sec. to get the answer for PDU. 2010-09-06 09:11:43,7, GSM1: Took 4 sec. to send a PDU. 2010-09-06 09:11:50,7, GSM1: Took 7 sec. to get the answer for PDU. 2010-09-06 09:11:56,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 09:11:58,7, GSM1: Took 2 sec. to get the answer for PDU. 2010-09-06 09:12:04,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 09:12:06,7, GSM1: Took 2 sec. to get the answer for PDU. 2010-09-06 09:12:12,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 09:12:19,7, GSM1: Took 7 sec. to get the answer for PDU. 2010-09-06 09:12:26,7, GSM1: Took 4 sec. to send a PDU. 2010-09-06 09:12:33,7, GSM1: Took 7 sec. to get the answer for PDU. 2010-09-06 09:12:41,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 09:12:43,7, GSM1: Took 2 sec. to get the answer for PDU.

It usually takes about five seconds to get the PDU sent, and after few seconds we get the answer which says that the message is sent.

With two modems the timing is different:

2010-09-06 10:06:06,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 10:06:07,7, GSM1: Took 1 sec. to get the answer for PDU. 2010-09-06 10:06:14,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 10:06:21,7, GSM1: Took 7 sec. to get the answer for PDU. 2010-09-06 10:06:28,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 10:06:54,7, GSM1: Took 26 sec. to get the answer for PDU. 2010-09-06 10:07:00,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 10:07:02,7, GSM1: Took 2 sec. to get the answer for PDU. 2010-09-06 10:07:08,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 10:07:38,7, GSM1: Took 30 sec. to get the answer for PDU. 2010-09-06 10:07:59,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 10:08:30,7, GSM1: Took 31 sec. to get the answer for PDU. 2010-09-06 10:08:49,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 10:09:02,7, GSM1: Took 13 sec. to get the answer for PDU. 2010-09-06 10:09:15,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 10:09:17,7, GSM1: Took 2 sec. to get the answer for PDU. 2010-09-06 10:09:24,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 10:09:50,7, GSM1: Took 26 sec. to get the answer for PDU. 2010-09-06 10:10:09,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 10:10:35,7, GSM1: Took 26 sec. to get the answer for PDU. 2010-09-06 10:10:55,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 10:11:02,7, GSM1: Took 7 sec. to get the answer for PDU. 2010-09-06 10:11:09,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 10:11:29,7, GSM1: Took 20 sec. to get the answer for PDU. 2010-09-06 10:11:35,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 10:11:37,7, GSM1: Took 2 sec. to get the answer for PDU. 2010-09-06 10:11:43,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 10:12:13,7, GSM1: Took 30 sec. to get the answer for PDU. 2010-09-06 10:12:35,7, GSM1: Took 6 sec. to send a PDU. 2010-09-06 10:13:00,7, GSM1: Took 25 sec. to get the answer for PDU. 2010-09-06 10:13:20,7, GSM1: Took 6 sec. to send a PDU. 2010-09-06 10:13:45,7, GSM1: Took 25 sec. to get the answer for PDU. 2010-09-06 10:13:57,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 10:14:22,7, GSM1: Took 25 sec. to get the answer for PDU. 2010-09-06 10:14:41,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 10:15:06,7, GSM1: Took 25 sec. to get the answer for PDU. 2010-09-06 10:15:13,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 10:15:14,7, GSM1: Took 1 sec. to get the answer for PDU. 2010-09-06 10:15:21,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 10:15:47,7, GSM1: Took 26 sec. to get the answer for PDU. 2010-09-06 10:16:07,7, GSM1: Took 4 sec. to send a PDU. 2010-09-06 10:16:33,7, GSM1: Took 26 sec. to get the answer for PDU. 2010-09-06 10:16:52,7, GSM1: Took 4 sec. to send a PDU. 2010-09-06 10:17:00,7, GSM1: Took 8 sec. to get the answer for PDU. 2010-09-06 10:17:06,7, GSM1: Took 5 sec. to send a PDU. 2010-09-06 10:17:08,7, GSM1: Took 2 sec. to get the answer for PDU.

and GSM2:

2010-09-06 10:05:55,7, GSM2: Took 5 sec. to send a PDU. 2010-09-06 10:06:15,7, GSM2: Took 20 sec. to get the answer for PDU. 2010-09-06 10:06:21,7, GSM2: Took 5 sec. to send a PDU. 2010-09-06 10:06:23,7, GSM2: Took 2 sec. to get the answer for PDU. 2010-09-06 10:06:29,7, GSM2: Took 5 sec. to send a PDU. 2010-09-06 10:06:37,7, GSM2: Took 8 sec. to get the answer for PDU. 2010-09-06 10:06:43,7, GSM2: Took 5 sec. to send a PDU. 2010-09-06 10:07:09,7, GSM2: Took 26 sec. to get the answer for PDU. 2010-09-06 10:07:16,7, GSM2: Took 4 sec. to send a PDU. 2010-09-06 10:07:23,7, GSM2: Took 7 sec. to get the answer for PDU. 2010-09-06 10:07:30,7, GSM2: Took 5 sec. to send a PDU. 2010-09-06 10:07:43,7, GSM2: Took 13 sec. to get the answer for PDU. 2010-09-06 10:07:49,7, GSM2: Took 4 sec. to send a PDU. 2010-09-06 10:07:52,7, GSM2: Took 3 sec. to get the answer for PDU. 2010-09-06 10:07:58,7, GSM2: Took 5 sec. to send a PDU. 2010-09-06 10:08:24,7, GSM2: Took 26 sec. to get the answer for PDU. 2010-09-06 10:08:44,7, GSM2: Took 5 sec. to send a PDU. 2010-09-06 10:09:10,7, GSM2: Took 26 sec. to get the answer for PDU. 2010-09-06 10:09:29,7, GSM2: Took 5 sec. to send a PDU. 2010-09-06 10:09:31,7, GSM2: Took 2 sec. to get the answer for PDU. 2010-09-06 10:09:42,7, GSM2: Took 5 sec. to send a PDU. 2010-09-06 10:09:44,7, GSM2: Took 2 sec. to get the answer for PDU. 2010-09-06 10:09:50,7, GSM2: Took 5 sec. to send a PDU. 2010-09-06 10:09:58,7, GSM2: Took 8 sec. to get the answer for PDU. 2010-09-06 10:10:05,7, GSM2: Took 5 sec. to send a PDU. 2010-09-06 10:10:07,7, GSM2: Took 2 sec. to get the answer for PDU. 2010-09-06 10:10:21,7, GSM2: Took 13 sec. to send a PDU. 2010-09-06 10:10:21,5, GSM2: Took 0 sec. to get the answer for PDU. 2010-09-06 10:10:27,7, GSM2: Took 5 sec. to send a PDU. 2010-09-06 10:10:29,7, GSM2: Took 2 sec. to get the answer for PDU. 2010-09-06 10:10:35,7, GSM2: Took 5 sec. to send a PDU. 2010-09-06 10:10:55,7, GSM2: Took 20 sec. to get the answer for PDU. 2010-09-06 10:11:05,7, GSM2: Took 5 sec. to send a PDU. 2010-09-06 10:11:13,7, GSM2: Took 8 sec. to get the answer for PDU. 2010-09-06 10:11:19,7, GSM2: Took 4 sec. to send a PDU. 2010-09-06 10:11:21,7, GSM2: Took 2 sec. to get the answer for PDU. 2010-09-06 10:11:28,7, GSM2: Took 5 sec. to send a PDU. 2010-09-06 10:11:53,7, GSM2: Took 25 sec. to get the answer for PDU. 2010-09-06 10:12:12,7, GSM2: Took 5 sec. to send a PDU. 2010-09-06 10:12:25,7, GSM2: Took 13 sec. to get the answer for PDU. 2010-09-06 10:12:33,7, GSM2: Took 6 sec. to send a PDU. 2010-09-06 10:12:35,7, GSM2: Took 2 sec. to get the answer for PDU. 2010-09-06 10:12:42,7, GSM2: Took 6 sec. to send a PDU. 2010-09-06 10:13:08,7, GSM2: Took 26 sec. to get the answer for PDU. 2010-09-06 10:13:28,7, GSM2: Took 5 sec. to send a PDU. 2010-09-06 10:13:48,7, GSM2: Took 20 sec. to get the answer for PDU.

It still takes about five seconds to get the PDU sent, but now there are longer delays while waiting for the response.

And when the delay is very long, about 25 seconds, result is CMS 500 and smsd will retry. This of course causes more delays and with single retry the sending time is about one minute. With two retries the time is about 100 - 120 seconds:

2010-09-06 09:52:40,6, GSM2: I have to send 1 short message for /var/spool/sms/checked/541xxx 2010-09-06 09:52:40,6, GSM2: Sending SMS from to 541xxx 2010-09-06 09:52:40,7, GSM2: -> AT+CMGS=152 2010-09-06 09:52:40,7, GSM2: Command is sent, waiting for the answer 2010-09-06 09:52:40,7, GSM2: <- > 2010-09-06 09:52:41,7, GSM2: -> PDU 2010-09-06 09:52:45,7, GSM2: Took 4 sec. to send a PDU. 2010-09-06 09:52:45,7, GSM2: Command is sent, waiting for the answer 2010-09-06 09:53:11,7, GSM2: Took 26 sec. to get the answer for PDU. 2010-09-06 09:53:11,7, GSM2: <- +CMS ERROR: 500 (Unknown error) 2010-09-06 09:53:11,3, GSM2: The modem answer was not OK: +CMS ERROR: 500 (Unknown error) 2010-09-06 09:53:11,5, GSM2: Waiting 10 sec. before retrying ... 2010-09-06 09:53:25,7, GSM2: -> AT+CMGS=152 2010-09-06 09:53:25,7, GSM2: Command is sent, waiting for the answer 2010-09-06 09:53:25,7, GSM2: <- > 2010-09-06 09:53:25,7, GSM2: -> PDU 2010-09-06 09:53:30,7, GSM2: Took 5 sec. to send a PDU. 2010-09-06 09:53:30,7, GSM2: Command is sent, waiting for the answer 2010-09-06 09:53:56,7, GSM2: Took 26 sec. to get the answer for PDU. 2010-09-06 09:53:56,7, GSM2: <- +CMS ERROR: 500 (Unknown error) 2010-09-06 09:53:56,3, GSM2: The modem answer was not OK: +CMS ERROR: 500 (Unknown error) 2010-09-06 09:53:56,5, GSM2: Waiting 10 sec. before retrying ... 2010-09-06 09:54:10,7, GSM2: -> AT+CMGS=152 2010-09-06 09:54:10,7, GSM2: Command is sent, waiting for the answer 2010-09-06 09:54:10,7, GSM2: <- > 2010-09-06 09:54:11,7, GSM2: -> PDU 2010-09-06 09:54:15,7, GSM2: Took 4 sec. to send a PDU. 2010-09-06 09:54:15,7, GSM2: Command is sent, waiting for the answer 2010-09-06 09:54:35,7, GSM2: Took 20 sec. to get the answer for PDU. 2010-09-06 09:54:35,7, GSM2: <- +CMGS: 179 OK 2010-09-06 09:54:35,5, GSM2: SMS sent, Message_id: 179, To: 541xxx, sending time 115 sec. Retries: 2

Those logs show that smsd does not cause this problem.

I think that it's still possible that this is caused by the problem in GSM network, but I have no tools to prove it. Later I will release a version of smsd for you, it has monitoring for GSM Cell ID codes and changes, but this does not tell why the modem has the delay.

You have that another PC which is sending with single modem, without problems at the same time when laptop is sending with another modem. Where is that PC located, how far away it is from the laptop?

With single stick on PC, and single stick on laptop, can you put both machines to send messages, and then move your laptop very near to the PC? I mean that sticks will be very close to each other. Does this make any difference?

You wrote that there was no alarms of power, but still it could be good to connect modems using a hub which has external power supply, at least 1000 mA rated. It may be possible that the alarm is missed, because GSM devices take more current when sending/communicating with network, and perhaps this peak is not measured by the laptop.

Keke :

In first place , thank you for your concern

I answer your questions:

1 - The laptop is about 2 metres from the PC. If both machines work with one modem each, (with identical smsd.conf), the system of the PC, seems to work better than the laptop. Of course, this situation become critical with two modems at the laptop (I can´t install two modems in the PC)

2 - In both machines, the systems start normally (without ERROR 500), for 15 or 20 minutes, after that , in both machines the errors appear and the speed comes down. It seems a GSM´s network problem.

3 - Tomorrow I test to move the laptop near the Pc, and see what happens. After that I have two more machines to work together (in different places in the office), and I will analyse the results of the test.

4 - At least, is possible the power problem. I check that with the different machines, and tell you what I find.

thanks again
Lextor

I was in the believe that with single modem both machines worked without errors, and error appeared only when you used two or more modems with laptop.

When unplugging a modem helps, the problem can be caused by the Huawei device. You could order BenQ M32 from Dealextreme and try if it works better.

But still make other checks and tests too.

That MTSMC-G-F4 which you had four months ago, it's not available anymore?

Keke:

No, unfortunatly no.

Today I try some modifications in SMSD conf file (for example senddelay =5), and the program runs much better in the PC but not in the laptop (with one modem each).

I´m going to test with a hub usb with power supply.

Lextor

Here is a new version of smsd for you (private), it will write logs with more information:

Hidden private text.

lextor wrote

Today I try some modifications in SMSD conf file (for example senddelay =5), and the program runs much better in the PC but not in the laptop (with one modem each).

How "better"? PDU sent faster, or less CMS errors, or longer time without errors after reboot?

I just checked from one of my systems: PDU is always sent in less than one second (no milliseconds was logged, but the "second digit" is always the same). Message is sent in 3 - 5 seconds.

Dear Keke

Less CMS errors, and with that the inactive time es lower.

Thanks with my "personal SMSD", I going to update with this version and send you two sepparate log, one with the pc, and one with the laptop. (in first place one modem each)

keke:

I have a little problem

I update the smsd version and run the application

I obtain this log:

18-09-10 12:47:29,2, smsd: Smsd v3.1.12 started.
18-09-10 12:47:29,2, smsd: Running as Administrador:Ninguno.
18-09-10 12:47:30,7, smsd: Running startup_check (shell): /tmp/smsd_script.ihuaGG /tmp/smsd_data.24EnYi
18-09-10 12:47:30,7, smsd: Done: startup_check (shell), execution time 0 sec., status: 0 (0)
18-09-10 12:47:30,4, smsd: File mode creation mask: 022 (0644, rw-r--r--).
18-09-10 12:47:30,5, GSM1: Modem handler 0 has started. PID: 2432.
18-09-10 12:47:30,3, GSM1: Couldn't open serial port /dev/com5, error: Device or resource busy, waiting 30 sec.
18-09-10 12:47:30,5, smsd: Outgoing file checker has started. PID: 392.
18-09-10 12:47:30,7, smsd: All PID's: 392,2432
18-09-10 12:47:35,2, smsd: Conflict with .LOCK file in the spooler: /var/spool/sms/outgoing/541161122925.txt /var/spool /sms/c hecked
18-09-10 12:47:35,2, smsd: Fatal error: Cannot move /var/spool/sms/outgoing/541161122925.txt to /var/spool/sms/checked. Check file and dir permissions.
18-09-10 12:47:35,2, smsd: Smsd mainprocess terminated abnormally. PID: 392.
18-09-10 12:47:35,3, GSM1: Cannot open serial port /dev/com5, error: Device or resource busy
18-09-10 12:47:35,2, GSM1: Modem handler 0 terminated. PID: 2432, was started 10-09-18 12:47:30.


It seems something abnormall, but the problem is the messages still "desapear" of the checked file- ¿where they go?

Lextor

lextor wrote

I update the smsd version and run the application

I obtain this log:

18-09-10 12:47:29,2, smsd: Smsd v3.1.12 started.

It's still 3.1.12, not your "personal smsd". Did you check which smsd and copy the new executable to the right place?

lextor wrote

It seems something abnormall, but the problem is the messages still "desapear" of the checked file- ¿where they go?

I think that you had smsd running when started a new one. The second smsd cannot open the port because it's in use. When a xxx.LOCK file exists in the checked directory, it causes the "conflict". Messages disappear because they are sent. Check the smsd.log and ps -ef listing.

You could use the script sms3 to start and stop the smsd, if not already using it.

Dear Keke:

I had a little problem with the upgrade of SMSD , because an error of compilation ( I need another package of Cygwin)

After that , I updated the application and run it in both machines.

I´d uploaded to you three files. one log "lp one modem" (laptop), one log "pc one modem", and one log "lp two modems"

I´m very lost about the problem. In the first minutes seems work very well, but after a few minutes I get the errors again.

It seems like the modems need a "reset" time to time.

I don´t know if is it a network problem, because at the same hour, the two machines works very different.

One question : ¿this version of SMSD get only more information or have anything else?

Thank you again. Lextor

There should not be any errors in compilation.

Do you have any other modems available (different than Huawei)?

Have you considered this?:

keke wrote

When unplugging a modem helps, the problem can be caused by the Huawei device. You could order BenQ M32 from Dealextreme and try if it works better.

The version of smsd has only changes which were listed.

You could take your laptop and move it to another address, which is at least one kilometer away from your office. Then test if the sending is faster.