Jan told me about this idea and we discussed on direct message channel.

So what we want to do is:

• Track the success rate of client
• If there is a client which has a rather bad success rate, we should deploy detailed telemetry for that client.
• If all clients have similar success rates, it might imply the problem for temporary storage initialization success rate is something else.

It seems the new telemetry would be added to InitializeOrigin

This needs to be tracked once per persistent type, per client. So the key for this might look like ${PersistentOrTemporary}${Client}

We shouldn't recalculate that for an origin twice (but for sure we won't submit origin as a key as a telemetry probe)

By this probe, we can also have a ballpark figure for the client usage.

I will do this next week.

The bug assignee didn't login in Bugzilla in the last 7 months.
:jstutte, could you have a look please?
For more information, please visit auto_nag documentation.

This isn't a priority right now.