# volume of "deletion-request" pings with `at_init` reason far exceeding `set_upload_enabled` reason If a user disables upload, we send a "deletion-request" ping. If that disabling happens between when the Glean SDK is init and when it is shutdown, that ping has a reason `set_upload_enabled`. If that disabling happens outside of that time, the ping is sent on the next Glean SDK init with reason `at_init`. We believe real human user actions can result in pings of either type. If they open the options and disable telemetry upload during browser startup, they can send an `at_init` ping. Glean is init [at some convenient time after the busiest parts of startup](https://searchfox.org/mozilla-central/rev/893a8f062ec6144c84403fbfb0a57234418b89cf/browser/components/BrowserGlue.sys.mjs#2713-2720), so especially on slower machines where Firefox may be in the busy part of startup for longer, users may have plenty of time to turn off upload. But more often we suspect this to be a reason associated with automation. If upload is opted out of through manipulation of the profile, then on the next (perhaps first) launch of Firefox with that profile, we'll upload an `at_init`-reason "deletion-request" ping. As such I'd expect the ratio of `at_init`-reason "deletion-request" pings to `set_upload_enabled`-reason "deletion-request" pings to be at most 1. [I'm wrong, of course](https://sql.telemetry.mozilla.org/queries/89640/source#221924). Even the ratio of clients sending `at_init` to `set_upload_enabled` regularly far exceeds 1 outside of nightly and aurora channels. This shouldn't have surprised me. I've seen this before in bug 1701985. In my defence, that was nearly 2 years ago. Conclusion: I was wrong to think that `set_upload_enabled` volume in clients or pings would exceed `at_init` volume in Firefox Desktop. Whoops. (( Unanswered questions remaining: those weird spikes starting Jan 5 ))
Bug 1809900 Comment 3 Edit History
Note: The actual edited comment in the bug view page will always show the original commenter’s name and original timestamp.
### volume of "deletion-request" pings with `at_init` reason far exceeding `set_upload_enabled` reason If a user disables upload, we send a "deletion-request" ping. If that disabling happens between when the Glean SDK is init and when it is shutdown, that ping has a reason `set_upload_enabled`. If that disabling happens outside of that time, the ping is sent on the next Glean SDK init with reason `at_init`. We believe real human user actions can result in pings of either type. If they open the options and disable telemetry upload during browser startup, they can send an `at_init` ping. Glean is init [at some convenient time after the busiest parts of startup](https://searchfox.org/mozilla-central/rev/893a8f062ec6144c84403fbfb0a57234418b89cf/browser/components/BrowserGlue.sys.mjs#2713-2720), so especially on slower machines where Firefox may be in the busy part of startup for longer, users may have plenty of time to turn off upload. But more often we suspect this to be a reason associated with automation. If upload is opted out of through manipulation of the profile, then on the next (perhaps first) launch of Firefox with that profile, we'll upload an `at_init`-reason "deletion-request" ping. As such I'd expect the ratio of `at_init`-reason "deletion-request" pings to `set_upload_enabled`-reason "deletion-request" pings to be at most 1. [I'm wrong, of course](https://sql.telemetry.mozilla.org/queries/89640/source#221924). Even the ratio of clients sending `at_init` to `set_upload_enabled` regularly far exceeds 1 outside of nightly and aurora channels. This shouldn't have surprised me. I've seen this before in bug 1701985. In my defence, that was nearly 2 years ago. Conclusion: I was wrong to think that `set_upload_enabled` volume in clients or pings would exceed `at_init` volume in Firefox Desktop. Whoops. (( Unanswered questions remaining: those weird spikes starting Jan 5 ))