We're tracking a server side issue that is probably having a major role in this OTA flakiness for some users. Stay tuned for more updates.
what is the status now?
i experienced the same - failed ota update
it was ok just a week ago
looking forward a solution soon
Start Uploading sketch GrowTower2_sep25a Over-the-Air
Binary not received by the board: board not found
Using library SPI at version 1.0 in folder: /home/builder/.arduino15/packages/arduino/hardware/samd/1.8.11/libraries/SPI [info] Using library adafruit_unified_sensor_1_1_4 at version 1.1.4 in folder: /home/builder/opt/libraries/latest/adafruit_unified_sensor_1_1_4 [info] Using library adafruit_bme680_library_2_0_1 at version 2.0.1 in folder: /home/builder/opt/libraries/latest/adafruit_bme680_library_2_0_1 [info] Using library adafruit_busio_1_9_1 at version 1.9.1 in folder: /home/builder/opt/libraries/latest/adafruit_busio_1_9_1 [info] Using library arduinoeccx08_1_3_5 at version 1.3.5 in folder: /home/builder/opt/libraries/latest/arduinoeccx08_1_3_5 [info] Using library SNU at version 1.0.2 in folder: /home/builder/.arduino15/packages/arduino/hardware/samd/1.8.11/lib
Any update yet??
I have exactly the same problem with mine since a couple of weeks !
Hope you'll find the problem, because it become to disappoint me...
3 posts were split to a new topic: Board identity changed after uploading via Arduino Cloud
Any updates for the OTA flakiness.
The IOT is useless if the board (nano 33 IOT) could not be loaded via OTA
Please treat this issues urgently
after several succesfull OTA scrips uploads on tuesday, I was realy happy , unfortunately today the problem is again.
I think that at night, the lower the traffic the easier to the cloud to synchronize with the remote boards.
just to give you some feedback
Hi all, I want to give you a quick update on the status of this issue.
The Cloud team is working on this topic: we are experimenting with our internal dev environment to replicate the issue and find a solution to make the OTA process reliable.
I know you are frustrated. I am frustrated, too, by not having already fixed this issue for all users of our platform.
It's hard to grok because, even if it often happens to some of you, it's not systematic and depends on many factors (load, events timing).
I'll update this thread as soon as possible.
I appreciate your patience.
i hope to hear some good news on the OTA issue
We completed the update to the MQTT broker: it now is using a new software version and a brand new sync protocol.
Can you try to run some OTA updates on your devices again and report here your success rates?
first test: Dashboard is working, Device is online, but
I'll keep trying OTA uploads with the other devices.
Thank you @Ale-X, let me know if it's improving with other devices or attempts: the success rate is improved a lot in our tests.
Now it' not working at all... : Unable to verify OTA: the board didn't answered after reboot
and the sketch is not updated at all even if the board actually has reboot
several test same sketch same board, same and different network connection used to upload the sketch to the board, connected always at his usual network
maybe all of us are overloading the ota traffic testing at the same time.
Here some lines of serial output from the board after doing a serial usb upload instead of OTA:
ArduinoIoTCloudTCP::handle_ConnectMqttBroker could not connect to mqtts-sa.iot.arduino.cc:8883
ArduinoIoTCloudTCP::handle_ConnectMqttBroker 2 connection attempt at tick time 33387
ArduinoIoTCloudTCP::handle_SubscribeMqttTopics could not subscribe to /a/t/ "...id#...."
Check your thing configuration, and press the reset button on your board.
ArduinoIoTCloudTCP::handle_SubscribeMqttTopics MQTT client connection lost
hope this could help to solve the proble as this is now a really unreliable situation
having some devices to make sketch upgrade OTA.
Thanks for the report! These failed connection attempts confirm a hypothesis we were working on this morning.
We just deployed more detailed logs and metrics to trace better when and how OTA requests go wrong.
@Ale-X could you please share a complete serial output from the board (before, during and after the OTA update attempt)?
Hi Zmoog I think I'm not able to have no serial output while doing OTA
I'll try a double connection one for uploading OTA et serial USB output
Just to share my experience. I disconnected one of my Nano 33 IOT which was connected OTA with a Thing sketch. I amended the sketch managed re-upload via OTA, WALA it's work. I then disconnected the device (power) leave it over night and connected it the next morning. WALA it's work. I hope the situation will remain as it is or get better with future software upgrade.
I've made one test yesterday afternoon ( 6:45 pm France time) with this scenario:
same board and sketch as before but a dual connection
A) Arduino cloud full editor to make the OTA upload to this board
B) Arduino IDE to have the USB serial output of the same board (also powered by its own 5V power supply (battery and module)
the OTA process started with some normal messages that went away as the board was restarted (a couple of times?) by the OTA and then it restarted normally with two Wi-Fi retries and one cloud successful attempt, I was in a hurry, so no time to save the serial output but nothing was abnormal as the OTA process was OK , I've leave the board running and since that I've made five OTA uploads to different boards without any problem...
So there is something you have modified yesterday that seems to have improved the reliability of the OTA, or the network traffic is really too much on workdays to keep synchro between devices and the cloud while answering from and to devices.
The only unusual thing I 've noticed is the double reboot of the board. that I've only noticed by the messages in the Serial output and several plug-in plug-out sounds from mi laptop
(but I think it's normal because the double verification and real write process of the OTA.)
I've no boards with the same setup here at home to repeat the experience and actually since yesterday all my OTA uploads have succeeded
Hope this could help to solve the issue or improve the reliability of the OTA as IMHO it's really one of the most important advantages of IOT projects.
I'll keep you informed when testing again