This thread is to discuss on the following device behaviour:
Photon(P1) device with Particle disconnected,but with WiFi connected.
(Background)
We have series of devices with the above observation. These devices are expected to pump data from connected sensors to a server over HTTP every minute. However, the behaviour is not as expected. It sends data at irregular intervals,sometimes with a gap of ~2-3 hours, resulting in tremendous data loss. Apart from that, device is also disconnected from cloud. Devices are not directly accessible as they are already deployed.
I would like to initiate a discussion and get opinion from experts on the possible root causes of the above behaviour before I jump to debugging the software.
(I am new to Particle system and just started working on this already deployed software.)
I can’t quite understand your problem description.
Does your topic title state what you intend to have or what your devices do, but you are not happy with?
If you want your devices disconnected from cloud but WiFi connected, the first info you’d need to provide:
What SYSTEM_MODE() do you use and do you have SYSTEM_THREAD(ENABLED)?
Also what system firmware version have you got on your devices?
On the other hand, if you see the devices still internet connected but not to the cloud and still have issues to reach your HTTP server, it might suggest that you have chosen the wrong SYSTEM_MODE() and/or inadequate actions to reconnect after connection loss.
Possibe root causes for any connection issues are so different and many that your description needs to provide some more detail in order to allow for focused analysis and feedback.
Does your topic title state what you intend to have or what your devices do, but you are not happy with?
This is the current state of the device, not the intended state. The intended device state is Particle connected and WiFi connected.
What SYSTEM_MODE() do you use and do you have SYSTEM_THREAD(ENABLED)?
It is SYSTEM_THREAD(ENABLED)
Also what system firmware version have you got on your devices?
0.4.6
On the other hand, if you see the devices still internet connected but
not to the cloud and still have issues to reach your HTTP server, it might suggest that you have chosen the wrong SYSTEM_MODE() and/or inadequate actions to reconnect after connection loss.
If the system mode is SYSTEM_THREAD(ENABLED), does the application code need to take specific actions to reconnect?
Possibe root causes for any connection issues are so different and many that your description needs to provide some more detail in order to allow for focused analysis and feedback.
I would be happy to provide all the asked details, could you please specify?
(About device) This is a flowmeter, intended operation is to capture the flowmeter parameters every minute and send to server. If transmission is not successful, save the data to SD card.
Have you got any way to upgrade the system, since 0.4.9 has got some major improvements incorporated on the whole radio/networking side.
Pretty much the opposite.
If you have SYSTEM_THREAD(ENABLED) (in conjunction with default SYSTEM_MODE(AUTOMATIC) - hence my question abour mode too), you should not try to reconnect manually otherwise you might get into your own way.
But you need to be aware that the cloud/network state can change anytime asynchronously to your application code and hence adding extra checks might be required before doing stuff that relies on active connections.
Have you got any way to upgrade the system, since 0.4.9 has got some major improvements incorporated on the whole radio/networking side.
Not yet. We would be moving towards 0.4.9 and the decision has already been made. At the same time, it is also important to diagnose these existing issues to avoid further occurrences in 0.4.9. Before flashing the new firmware, we need to pin point these issues.
Pretty much the opposite.
If you have SYSTEM_THREAD(ENABLED) (in conjunction with default SYSTEM_MODE(AUTOMATIC) - hence my question abour mode too), you should not try to reconnect manually otherwise you might get into your own way.
Ok.
But you need to be aware that the cloud/network state can change anytime asynchronously to your application code and hence adding extra checks might be required before doing stuff that relies on active connections.
The checks are already present (WiFi.ready() and Particle.connected()) at specific points. The issue still exists.
I am looking for the possible reasons for this issue from System firmware perspective. If you can suggest some check points, I don't mind debugging the firmware immediately.