Thank you everyone for your patience.
Our team has put in place an intermediary workaround - this is to ensure that APs do not reboot en-masse outside of expected hours, causing business disruption. The details of the workaround are as follows:
1. If an AP has an uptime of <2 hours (e.g. recently rebooted or coming online for the first time) and is stalled in this upgrade pending state, it will complete its upgrade 0-2 hours after coming online.
2. If an AP has an upgrade scheduled from here on, the device will upgrade 0-2 hours from the scheduled upgrade time.
3. Otherwise, if an AP does not fall into the above categories (e.g. it first came online after more than 2 hours uptime) it will upgrade 1-2 hours after the network's next scheduled maintenance window.
You can read more about how to configure the maintenance window here.
We appreciate your patience whilst we work through this issue with care. We will be able to advise once we've fully resolved the issue soon.
Whilst I am a Meraki employee, some of what I post may be opinion (especially architecture!). Others may have better or more efficient ways of doing things, so please learn from everyone!