BGS2T - mc55I FTP Download problem | Thales IoT Developer Community
April 10, 2019 - 11:26am, 2209 views
hi
I use BGS2T (mc55i in some older system) in order to download update sotfware from a FTP server.
It worked for a year.
In the last 20 days ha stopped working.
In the downloaded file I find different gaps (2 -3 KBytes in a file of 1.5MByte) but no error is reported by the ME.
What could be the problem?
Hello,
Could you give an example how does such 'gap' look like? Is the file content corrupted or is additional data added? Have you changed anything in your application or module's firmware? Or maybe something has changed on the server?
Does the problem occur in all devices or just one? Please paste ATI1 output.
Regards,
Bartłomiej
hi
http://www.genovadanza.com/tmp/GapData.zip
In this zip I've packed the complete file and the downloaded file. They are in base64 format.
Data seems only missing.
I haven't changed any firmware.
I've tried with two different FTP server.
This appears to occur in every device.
Hello,
The link is empty. Could you provide AT commands log to show how FTP client is configured and file downloaded? Are AT commands used for that?
Please also paste ATI1 output.
What device is controlling BGS2 terminal - is it possible that the data is being corrupted on that device after download?
Regards,
Bartłomiej
Hi
This is the correct link. http://www.genovadanza.com/tmp/GapData.zip
I'm searching a sniffer RS232 when I'll find I'll send the complete AT sequence.
I use a board with WinCE.
I load data and save them directly in a file. However the same software one month ago worked well
Regards
Massimiliano Zirilli
Hi
here you can find the comunication's log
http://www.genovadanza.com/tmp/LogSeriale.txt
In this example it jumped almost 5kbyte
Hello,
As I understand the bigger file (CU200_FW_OFFICIALa.b64) is the original file being downloaded from the server with the complete contents and the smaller one (CU200_FW_OFFICIAL.b64) is what has been stored in WinCE device as the effect of the download.
I have compared the both files contents and I have found two places where a part of the data is missing in the downloaded file.
Next I copied a few data sequences from the missing fragments and was able to find all these fragments in the AT commands log. I have only tried with some data from the beginning, middle and end of the missing blocks; not every byte.
So if the AT commands log and the downloaded file were both captured duirng the same download operation, it seems that probably all the data has been downloaded and displayed by the module's AT interface and your WinCE device has received it. But for some reason not all was stored to the file. I think that you need to verify the software on WinCE now.
Best regards,
Bartłomiej
Hi
Unfortunately the log is referred to a following download :-(.
Data are not present. The sum of the AT^SISR values it's exactly the same of the downloaded file that is always smaller than the original.
Best regards
Hello,
If you sent the AT commands log and downloaded file from the same session, it would be then possible to compare and find some dependencies.
If the total amount of data confirmed by the module with ^SISR: commands is equal to the amount that was stored to the file and less then in the original file, it looks like some of the data was not received from the network. Getting pcap trace from the module could be helpful but you can't do it. Does it also happen on MC55i? Maybe you should try to test with some other operator then.
Regards,
Bartłomiej