I do see the dump files now that you mentioned it use the web socket console. There are 8 of them total. I'll try and get one pushed to you later today when I get the device.

We're now using SDK 4.2.0-8775.  Should I upgrade to be consistent?

Now that you mention it, it might be some of our code causing the core dumps.  The core functionality is using MQTT code connecting to AWS, Temperature sensors (by TI), and sending GPIO functions to certain pins for functionality on our side.

Right now, I don't have the temperature sensors hooked up, so that could be the reason for the crashes.

In the meantime, are there any settings for safemode to bring it back online / network connected when it hits this "safe mode" state?

Just wondering if there are options or issues when we start shipping some of these development boards to remote locations.

Thanks for all of the help.

I have a full development WGM160P board and we do pushes to it mostly via Gecko OS IDE and then through Zentri updates using OTA enabled settings.  We then call a dfu_update on the device after we release a bundle.

For some reason we are getting the device into a weird state after it's running.  Here's the error.

SAVEMODE> Starting Safemode console for 300 seconds.
Issue 'faults_reset' then 'reboot' to exit safemode.

So far, after that we must connect it to a PC with USB.

Is this only because we are pushing via Gecko OS IDE?

after I run those two commands (connected via computer), the device resets fine and runs our code and connects to the internet fine.

We're wanting to test some things completely unattended and always run Zentri commands or is this just an issue with just using Development Builds?  I haven't tried a Production Build via Zentri.

the system.uuid is:   EADE2FF3BF69FCAD93D8518ECCCCCCFFFE2F88D1

I was going to test with just using Gecko OS Studio -> Program Device -> Mode -> Products -> <Select Product> only, but not sure if that's a better solution to prevent the issue.

I'm using the WGM160P wireless board and were using Gecko Studio IDE with it.  Now I'm working fine on a Mac, but  when I move the code and run it on Windows, I get the following error.

11:33:14 **** Incremental Build of configuration Standard Build for project FirstMateControlsHub ****
"C:\\Users\\npneu\\FirstmateControlsHub\\FirstMateControlsHub\\make.sh" all 
Cannot run program "C:\Users\npneu\FirstmateControlsHub\FirstMateControlsHub\make.sh": Launching failed

Error: Program "C:\Users\npneu\FirstmateControlsHub\FirstMateControlsHub\make.sh" not found in PATH
PATH=[C:/Users/npneu/AppData/Roaming/Gecko OS Studio/tools/ide/GeckoOSIDE_x64//features/com.silabs.external.java.windows.x86_64.feature_1.8.0.92/jre/bin/server;C:/Users/npneu/AppData/Roaming/Gecko OS Studio/tools/ide/GeckoOSIDE_x64//features/com.silabs.external.java.windows.x86_64.feature_1.8.0.92/jre/bin;C:/Users/npneu/AppData/Roaming/Gecko OS Studio/tools/ide/GeckoOSIDE_x64//features/com.silabs.external.java.windows.x86_64.feature_1.8.0.92/jre/lib/amd64;C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOWS\System32\Wbem;C:\WINDOWS\System32\WindowsPowerShell\v1.0\;C:\WINDOWS\System32\OpenSSH\;C:\Program Files (x86)\Intel\Intel(R) Management Engine Components\DAL;C:\Program Files\Intel\Intel(R) Management Engine Components\DAL;C:\Program Files (x86)\Intel\Intel(R) Management Engine Components\IPT;C:\Program Files\Intel\Intel(R) Management Engine Components\IPT;C:\Program Files\Git\cmd;C:\Users\npneu\AppData\Local\Microsoft\WindowsApps;;C:\Users\npneu\AppData\Roaming\Gecko OS Studio\tools\ide\GeckoOSIDE_x64;]

11:33:14 Build Finished (took 99ms)

Is this somewhere I won't be able to use code easily between both a Mac and a Windows PC?  I searched the forums and didn't see much on how to fix this issue.

I have two development WGM169P boards and I have one that will fail on connecting to a few wifi networks, while the other one will not.  This is causing problems with the network_verify call.  I will either get a Timeout or a Command Failed in the REST response.  I also have tried resetting the board using the Gecko OS Studio by using the Program Device / Mode "Default SILABS-WGM160P) Program button.  I'll try the setup_web function and go into the web version and it will fail as well.

I output and preserved both "get all" data between both boards and I did a compare between both and didn't even see much differences between the two other than mac address type of settings.

Here's the rough command (wrong hash for privacy).

{
   "flags"   : 0,
   "command" : "network_verify wifi Nickernet_Guest 8A:96:4E:A9:9A:D1 11 4 12345678901234567890123456789012345678901234567890"
}

Maybe I'm not resetting the board properly?  It doesn't make much sense because I've compared the request data between both boards and they're the same.

Any thoughts?

After digging a bit more and determining it wasn't client side code, I fixed the problem with the following:

http.server.keep_alive_enabled 1

https://docs.silabs.com/gecko-os/4/standard/latest/cmd/variables/http-server#http-server-keep-alive-enabled

I added this to my startup script and things are working fine.

set, http.server.keep_alive_enabled 1, \r\n#Enabling Keep Alive to reduce the retry issues on several commands

I kept the default http.server.keep_alive_timeout

So far, I haven't seen any failures on iOS applications.  I haven't tried Android yet, but this appears to have fixed the issue for me.

Could it have anything to do with calling by IP Address.  I found this code in connect.js.  I'm not using the web_setup, but trying to use it as a guide to help fix this issue potentially..

Does it have anything to do with this code using mdns?  I'm not sure if the web setup code is using Ip address or mdns for the URL yet..  I'm also not using mdns yet.

if(self.device.get('web_setup')) {
      cmds.push({cmd: 'set', args: {args:'wl o e 1'}});

      if(self.reconnect){
        var mdns = $(this.el).find('input[name="mdns"]').val().trim();
        var mac = self.device.get('mac');
        mac = mac.substring(mac.length - 4).replace(':','');

        mdns = mdns.length > 0 ? mdns : 'gecko_os-' + mac;
        mdns = mdns.toLowerCase();

        self.mdns = mdns;

        cmds.push({cmd: 'set', args: {args:'ht s e 1'}});
        cmds.push({cmd: 'set', args: {args:'md e 1'}});
        cmds.push({cmd: 'set', args: {args:'md n ' + mdns}});
        cmds.push({cmd: 'set', args: {args:'md s http'}});
        cmds.push({cmd: 'set', args: {args:'ht s c *'}});
      }
    }

I noticed in the web setup code that the timeout value is set really high upwards of 20,000 or 20 seconds on the network call.

    self.device.issueCommand(
      {property: 'mdns', cmd: 'get', args: {args: 'md n', ret: false}},
      function(err, res) {
        self.device.geckoOS.scan({args: '-v', timeout: 20000}, scanComplete);
      });