I have a webhook integration that works fine usually. Today while analyzing data from Sept 5 to present, I noticed that one of our Boron devices was publishing a fw_version key value of 0 for several days, and then started publishing the correct fw_version of 10623 beginning 3pm UTC on Sept 7th.
Based on the content of these webhooks that we have saved on our sever, it looks like the firmware version that this Boron was running was actually 10623 during this time frame.
Note that this firmware version is running on 6 other Borons in our product, and they are all reporting fw_version of 10623 throughout the entire period under investigation.
What could cause the particle servers to set fw_version = 0 in a particle product webhook integration when it is in fact non-zero?