Greengrass v2 fleet provisioning retry in case of lack of the Internet connection.

0

Hi.

I have created a Greengrass fleet deployment workflow for my devices which works great - cloud resources are provisioned, claim certs are used to obtain device individual certs, and device gets connected. However in production environment the scenario will be different:

  1. An IoT device is installed by an engineer in a newly built home, with no occupiers and no internet connection.
  2. The Greengrass is ready to claim device certs and connect as soon as the device is powered up.
  3. The device then is powered up by the installing engineer, meaning the Greengrass fleet provisioning plugin will try to do its job, but there is no internet connection, so it will fail.
  4. The engineer will leave the device on, and it will be permanently on since now on.
  5. The occupiers will move in some time later and install their broadband router, the Internet is now available.

At this point, from my tests, I noticed the Greengrass/provisioning plugin won't detect Internet connection and won't try to register the device/obtain certs - meaning there is no retry functionality implemented.

I understand that the device can be rebooted or Greengrass service restarted to initialize new registration attempt, however - if possible I'd like to avoid occupiers fiddling with the device, and prefer the device to the retry automatically.

I also understand that I can write a software that will be automatically cyclically restarting the Greengrass service if the previous registration attempt failed (probably by checking if the thingCert.crt is present) but before I'll spend my time on coding, the question is:

Is it possible to configure Greengrass to retry delayed provisioning using claim certificates if previous attempt has failed?

Appreciate any help.

  • Hi K676, can you please provide the exact error which is being logged? The error cause should be MqttException and I'd like to know the precise error message included with that exception.

K676
已提問 2 年前檢視次數 364 次
1 個回答
1
已接受的答案

Hello,

Unfortunately, we do not support configurable provisioning retry, which is a gap in our current set up and we will address in future. In the meantime, I would recommend you go ahead with your workaround.

AWS
已回答 2 年前

您尚未登入。 登入 去張貼答案。

一個好的回答可以清楚地回答問題並提供建設性的意見回饋,同時有助於提問者的專業成長。

回答問題指南