For clients sending "default" reason in New Profile Pings, Chutten's intuition that most of these are coming from users sending a second New Profile Ping on a later session is correct. Most of the enormous spike in "default" reason New Profile Pings in 2023 (see scary blue line here: https://sql.telemetry.mozilla.org/queries/101818/source#250852) are extraneous pings coming from profiles that have *already* sent us a New Profile Ping with the correct startup reason (firstrun, etc.). I don't think we need to worry too much about this group, since the New Profile definition in our downstream datasets de-dupes these. There *is* a population of users who send us a "default" reason New Profile Ping as their first New Profile Pings. Before August 2022, this group was about 5-6% of new profile ping clients (deduped). After August 2022, we see a big spike (incident?) and then the level shifts to about 10%. That still seems like a lot. Would "Firefox crashed before NPP got sent" be common enough to explain 10% (of daily new client ids that send us a New Profile Ping that day)?
Bug 1911014 Comment 14 Edit History
Note: The actual edited comment in the bug view page will always show the original commenter’s name and original timestamp.
For clients sending "default" reason in New Profile Pings, Chutten's intuition that most of these are coming from users sending a second New Profile Ping on a later session is correct. Most of the enormous spike in "default" reason New Profile Pings in 2023 (see scary blue line here: https://sql.telemetry.mozilla.org/queries/101818/source#250852) are extraneous pings coming from profiles that have *already* sent us a New Profile Ping with the correct startup reason (firstrun, etc.). I don't think we need to worry too much about this group, since the New Profile definition in our downstream datasets de-dupes these. There *is* a population of users who send us a "default" reason New Profile Ping as their first New Profile Pings. Before August 2022, this group was about 5-6% of new profile ping clients (deduped). After August 2022, we see a big spike (incident?) and then the level shifts to about 10%. * see: https://sql.telemetry.mozilla.org/queries/101849/source#250926 That still seems like a lot. Would "Firefox crashed before NPP got sent" be common enough to explain 10% (of daily new client ids that send us a New Profile Ping that day)?
For clients sending "default" reason in New Profile Pings, Chutten's intuition that most of these are coming from users sending a second New Profile Ping on a later session is correct. Most of the enormous spike in "default" reason New Profile Pings in 2023 (see scary blue line here: https://sql.telemetry.mozilla.org/queries/101818/source#250852) are extraneous pings coming from profiles that have *already* sent us a New Profile Ping with the correct startup reason (firstrun, etc.). I don't think we need to worry too much about this group, since the New Profile definition in our downstream datasets de-dupes these. There *is* a population of users who send us a "default" reason New Profile Ping as their first New Profile Pings. Before August 2022, this group was about 5-6% of new profile ping clients (deduped). After August 2022, we see a big spike (incident?) and then the level shifts to about 10%. * see: https://sql.telemetry.mozilla.org/queries/101828/source#250881 * see: https://sql.telemetry.mozilla.org/queries/101849/source#250926 That still seems like a lot. Would "Firefox crashed before NPP got sent" be common enough to explain 10% (of daily new client ids that send us a New Profile Ping that day)?