If transcodes fail due to some temporary problem -- like ffmpeg not being installed or a bug that later gets fixed -- it should be possible for users to re-start them in some way, even if it's limited.
Currently there doesn't seem to be a way to do this from the UI, and the jobs runner for transcoding only seems to handle new things in the queue, without giving a way to run specific files.
Version: unspecified
Severity: normal
mdale wrote:
thanks for the bug reports,
Do we want to limit this to admins? or expose this to general users in the case where the fail failed to transcode and some time has passed?
I imagined that when we update ffmpeg we would run a maintenance script that would clear out all the failed transcodes, and the value of exposing a retry button to end users would be of limited usefulness. ( i.e it would only help in cases where things went very very wrong on our back end ) Or worse a broken encode gets re-tried and re-tried by everyone because ffmpeg / our system thinks its successful but its really broken ... If the user wanted to retry they would do good to upload a different copy of the asset if they had one available ( once bug 29179 is fixed )
Maybe we do a middle ground and only expose this to privileged users of some kind? and if we find people need to do it a lot we can broaden its privilege exposure?
mdale wrote:
added this feature in r89973 Restricted to the 'transcode-reset' right. When viewing the assets page you can reset transcodes by making the associative api call.