Ethernet gateway troubleshooting advice

[hek]

The SOftSPI enable RF24 and DigitalIO library has been added to master a few days ago. You should probably not use "development" at the moment.

[niccodemi]

I noticed something was different. About 8 days ago I was able to get Uno r3 and w5100 shield work using development library, which has just been updated by anticimex. gateway ran for a week without issues except i was not able to control relay nodes. Yesterday I downloaded master and development libraries. With Master i can get gateway to communicate with vera but when I press include button radio comms stop. i also tried development library. modified as per dan and anticimex instructions and not modified. in both cases i can upload sketch but then i do not get any serial communication and also pinging gateway does not work.

[Dan S.]

@niccodemi I should mention that in the new master there is a comment in the gateway code that says:

"Remove UIPEthernet include below and include Ethernet.h." In other words the default is not Wiznet. If you don't change the default (which I didn't at first) then you will have a commo problem since it is for ENC28J60 shield.

[niccodemi]

@Dan-S. thanks but I have been aware of that. I will try to patch Master library as per your advice. Btw do you use any sensor node with your gateway? Do you use w5100 module or shield? Mine is same as
on this site

[Dan S.]

@niccodemi I am using a genuine arduino uno and arduino ethernet shield which is similar to the one you reference, but don't think that should make any difference if you are using soft spi. Currently I have a light sensor and a temp/humidity sensor reporting to the gateway and all are working fine. Plan to expand more now that it is working so dependably.

[karenbobiv]

@Dan-S. Thanks much. After a good night sleep I gave it another try. Made changes, reloaded script, and got the Gateway started message in the serial monitor. I can also ping the GW. However, in the "nothing is ever easy" category, now it won't include my temperature sensors. I attached the temperature sensors to the serial monitor and get the following:

sensor started, id 1
send: 1-1-0-0 s=255,c=0,t=17,pt=0,l=3,st=fail:1.4
send: 1-1-0-0 s=255,c=3,t=6,pt=1,l=1,st=fail:0
send: 1-1-0-0 s=255,c=3,t=11,pt=0,l=18,st=fail:Temperature Sensor
send: 1-1-0-0 s=255,c=3,t=12,pt=0,l=3,st=fail:1.0
send: 1-1-0-0 s=0,c=0,t=6,pt=0,l=3,st=fail:1.4
send: 1-1-0-0 s=1,c=0,t=6,pt=0,l=3,st=fail:1.4
send: 1-1-0-0 s=2,c=0,t=6,pt=0,l=3,st=fail:1.4
send: 1-1-255-255 s=255,c=3,t=7,pt=0,l=0,st=fail:
send: 1-1-0-0 s=0,c=1,t=0,pt=7,l=5,st=fail:74.3
send: 1-1-0-0 s=1,c=1,t=0,pt=7,l=5,st=fail:83.7
send: 1-1-0-0 s=2,c=1,t=0,pt=7,l=5,st=fail:73.0

Tried inclusion lots of times. Unplugged and plugged in power for sensor network when I did that to try to get the "handshake." Likely a comedy of errors on my part, but I can't see what's wrong now. Thanks.

[karenbobiv]

Unplugged the GW and plugged back in as part of effort to diagnose and now I can't ping it and it's not talking to plugin. I checked it with serial monitor and it still shows GW Startup. Not sure what to try now.

[Dan S.]

I always interpreted the GW startup message to mean that the radio part of the gateway is working. So you've made good progress. When you start it up and plug it into the network, what happens when you hit reload on Vera. If it reloads with no errors, I always took that to mean that the ethernet part is working and the gateway is talking to Vera. For push the include button on the vera plugin and for as long as the inclusion person lasts continue to start and restart the sensor until the plugin says it is detected. Sometimes takes a couple of tries. But your posting of the sensor messages says the gateway is not talking to the sensor. But first you need to get it included and it may take a few tries. Hope this helps.

[karenbobiv]

@Dan-S. Not sure why, but can ping again and talking with plugin. Interesting development. I noted that the sensor appeared to talk to GW (st=ok?) and when I saw that I tried to include, but then it would switch back to st=fail and wouldn't include. Did this several times. Building another simple temp sensor to try that.

send: 1-1-0-0 s=2,c=1,t=0,pt=7,l=5,st=fail:73.6
send: 1-1-0-0 s=0,c=1,t=0,pt=7,l=5,st=ok:73.6
send: 1-1-0-0 s=1,c=1,t=0,pt=7,l=5,st=ok:73.9
send: 1-1-0-0 s=1,c=1,t=0,pt=7,l=5,st=ok:73.8
send: 1-1-0-0 s=2,c=1,t=0,pt=7,l=5,st=ok:73.5
send: 1-1-0-0 s=2,c=1,t=0,pt=7,l=5,st=ok:73.6
send: 1-1-0-0 s=2,c=1,t=0,pt=7,l=5,st=fail:73.5
send: 1-1-0-0 s=0,c=1,t=0,pt=7,l=5,st=ok:73.5
send: 1-1-0-0 s=1,c=1,t=0,pt=7,l=5,st=ok:73.7
send: 1-1-0-0 s=1,c=1,t=0,pt=7,l=5,st=fail:73.8
send: 1-1-0-0 s=2,c=1,t=0,pt=7,l=5,st=fail:73.3
send: 1-1-0-0 s=1,c=1,t=0,pt=7,l=5,st=ok:73.7
send: 1-1-0-0 s=1,c=1,t=0,pt=7,l=5,st=ok:73.8
send: 1-1-0-0 s=2,c=1,t=0,pt=7,l=5,st=ok:73.5
send: 1-1-0-0 s=1,c=1,t=0,pt=7,l=5,st=ok:73.7

[karenbobiv]

I was able to add the Node, but no sensors show up. Do I need to reload the sketch for the temp sensor now that I switched to the soft SPI? Does that sketch change somehow? Tried to reload temp sketch on the node, but got this error. DigitalIO.h is in the library.

Arduino: 1.5.8 (Windows 7), Board: "Arduino Nano, ATmega328"

In file included from C:\Users\Robert\Documents\Arduino\libraries\MySensors/MySensor.h:28:0,
from DallasTemperatureSensor.ino:2:
C:\Users\Robert\Documents\Arduino\libraries\MySensors/utility/RF24.h:20:23: fatal error: DigitalIO.h: No such file or directory
#include <DigitalIO.h>
^
compilation terminated.
Error compiling.

[hek]

You should only enable SoftSPI when compiling the ethernet gateway. Change it back when this has been done and you want to build your sensors.

[niccodemi]

I used latest Master library this time and patched RF24_config.h (enabled SoftSPI and changed pins), included DigitalIO. Sketch uploaded successfully, I get "gateway startup" message and I can ping it. Vera also communicates with gateway (lib version shows up, inclusion buttons work). Then I tried to include two independent nodes / sensors : distance and relay. I can only include relay node (no light button) and distance sensor (no node). I tried several times. Gateway always receives radio message but somehow doesn't relay all info to Vera? I rebuilt sensors with latest Master library (with softspi disabled). What else should I try?

[hek]

@niccodemi

Could you post serial monitor debug messages from gateway (maybe easiest when plugged into computer). Then restart a sensor to get some traffic.

[niccodemi]

@hek see below, I included relay node, tried several times, only 1 device added (node, no relay button)

0;0;3;0;14;Gateway startup complete.
0;0;3;0;9;read: 3-3-0 s=1,c=1,t=13,pt=2,l=2:5
3;1;1;0;13;5
0;0;3;0;2;Get Version
0;0;3;0;2;1.4
0;0;3;0;9;read: 3-3-0 s=1,c=1,t=13,pt=2,l=2:21
3;1;1;0;13;21
0;0;3;0;5;1
0;0;3;0;5;1
0;0;3;0;9;read: 4-4-0 s=255,c=0,t=18,pt=0,l=3:1.4
4;255;0;0;18;1.4
0;0;3;0;9;read: 4-4-0 s=255,c=3,t=6,pt=1,l=1:0
4;255;3;0;6;0
0;0;3;0;9;read: 4-1-0 s=255,c=3,t=6,pt=1,l=1:0
4;255;3;0;6;0
0;0;3;0;9;read: 4-4-0 s=255,c=0,t=18,pt=0,l=3:1.4
4;255;0;0;18;1.4
0;0;3;0;9;read: 4-1-0 s=255,c=0,t=18,pt=0,l=3:1.4
4;255;0;0;18;1.4
0;0;3;0;9;read: 4-4-0 s=255,c=3,t=6,pt=1,l=1:0
4;255;3;0;6;0
0;0;3;0;9;read: 4-1-0 s=255,c=3,t=6,pt=1,l=1:0
4;255;3;0;6;0
0;0;3;0;5;0
0;0;3;0;5;0
0;0;3;0;2;Get Version
0;0;3;0;2;1.4
0;0;3;0;9;read: 4-4-0 s=255,c=0,t=18,pt=0,l=3:1.4
4;255;0;0;18;1.4
0;0;3;0;9;read: 4-1-0 s=255,c=0,t=18,pt=0,l=3:1.4
4;255;0;0;18;1.4
0;0;3;0;9;read: 4-4-0 s=255,c=3,t=6,pt=1,l=1:0
4;255;3;0;6;0
0;0;3;0;9;read: 4-1-0 s=255,c=3,t=6,pt=1,l=1:0
4;255;3;0;6;0
4;255;3;0;6;M
0;0;3;0;9;send: 0-0-1-4 s=255,c=3,t=6,pt=0,l=1,st=fail:M
4;255;3;0;6;M
0;0;3;0;9;send: 0-0-1-4 s=255,c=3,t=6,pt=0,l=1,st=fail:M
0;0;3;0;9;read: 4-4-0 s=255,c=3,t=11,pt=0,l=5:Relay
4;255;3;0;11;Relay
0;0;3;0;9;read: 4-4-0 s=255,c=3,t=12,pt=0,l=3:1.0
4;255;3;0;12;1.0
0;0;3;0;9;read: 4-4-0 s=1,c=0,t=3,pt=0,l=3:1.4
4;1;0;0;3;1.4

[hek]

Looks like the sends is failing.

@Anticimex told me he was investigating send-fails that evidently started when the updated RF24-with-SOFTSPI was introduced a couple of weeks ago.

Hmm... somehow this new version could be affecting this. If rolling back would mean you'd lose SOFT-SPI-option.

@Anticimex: Did you locate the problem?

[hek]

Something fishy is definitely going on: This is the log from a relay actuator I just hooked up.

send: 5-5-0-0 s=1,c=0,t=3,pt=0,l=3,st=ok:1.4
read: 5-1-0 s=1,c=0,t=3,pt=0,l=3:1.4
read: 5-1-0 s=1,c=0,t=3,pt=0,l=3:1.4
read: 5-1-0 s=1,c=0,t=3,pt=0,l=3:1.4

After the last send it finds messages (the last 3) which gateway not should have sent. Almost as the gateways hardware ack-messages is picked up by the sensor as regular messages.

You might wanna roll back until the root cause of the is found.

Pinging @TMRh20Projects here also.

This is the commit that is causing the problem: https://github.com/mysensors/Arduino/commit/2fd948fc6667460961194040bef635d25b4dfd8b

[hek]

I think I've found the problem and a patch has been check in to master branch. So @niccodemi download the latest and try again please.

@marceltrapman this might be related to some the problems you experiences earlier as well.

[marceltrapman]

@hek said:

@marceltrapman this might be related to some the problems you experiences earlier as well.

I would be really happy when my case is solved as well with this.
Do I have to reload gateway, repeater, sensor or all sketches?

[hek]

Yes, as it is related to acks both gateway and sensors is affected, But it does not solve the fetch-id-through-repeater problem we also discussed.
I really encourage people to update if they are running code from 10:th of september and after.

[Anticimex]

I have not yet had opportunity to dig. Is the problem already identified and solved?