This topic is just to report a little bug which confused me a lot though and made debugging a bit longer.
Use case: when Client can access the backend but NOT the file storage proxy (for example, outgoing traffic is blocked on port 9000 but allowed on 443 ), UI shows the following when such device is starting update:
which is misleading because in fact if you look at the device logs, download is not even starting.
The device stays in this state for a day retrying until finally reports a failed attempt.
Expected behavior: it should show “Starting Download” or at least “Downloading…0%” during this period.
Tested with mender clients V2.0, backend V2.1 and V2.3.0
This is another issue but also related to the Deployments Page:
Currently, “Started At” timestamp is always the same for all devices across a given deployment and corresponds to the time when deployment was created and NOT when it started for a given device:
(I’m sure that devices on this screenshot haven’t started the update at the same minute)
While it makes more sense, IMO, to show the timestamp of when the device started the update (i.e. when deployment status is switching from “pending” to “downloading” for a given device).
Otherwise, this column brings no useful information to the user, as the timestamp of the beginning of the whole deployment is already shown on deployment info box:
P.S. Is Mender Hub is the only way to report bugs/propositions?