-
Notifications
You must be signed in to change notification settings - Fork 116
feat(): stores update clients references to new profile clients #7720
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
base: main
Are you sure you want to change the base?
feat(): stores update clients references to new profile clients #7720
Conversation
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
…stats_v1 with new_profile_clients and update retention source
… new datatasets so that the other two can be deprecated
0e6cf1c
to
186963f
Compare
Integration report for "update firefox_android_clients and retention_week_4 references to use new datatasets so that the other two can be deprecated"
|
Would we be able to make these changes in a V2 table, that way you can backfill the V2 and compare with V1? Then once we are satisfied everything looks good and have run QA checks, we can swap the views to point at the new V2 tables and deprecate the V1 tables at that time? |
The iOS app campaign table is not needed right now. |
@soGaussian are you saying |
ok, I will create a follow-up ticket then to deprecate that dataset. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This has now been marked as deprecated.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Found some differences, dived into a specific client and it appears the new table is reporting activation field correctly. The old dataset not (as per the activation logic). Hard to say why exactly, but this could be because the underlaying dataset we rely on for calculating this changes somehow or potentially the definition of activation changed since when the activation information for that client was generated.
We only see differences for dates older than (including) 2024 August.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
ok, so historically there's definitely bigger differences. And seeing that the aggregation is generated at runtime using client level info so even historically I'd expect more alignment since both datasets should be Shredded so this should not be a factor for these differences. The only thing I can really think of is that since the differences between new_profile count and activation count show up starting August 2024 is that the definition of activation changes around then and this is the definition used by the new dataset (since we seem to have match since then).
Another theory would be that the new dataset was created and backfilled around the time we see differences and historical data differs because at the backfill time the underlaying data (used to create new profiles and activation datasets) may have changed which is why historically we see differences and going forward we do not as both datasets would be updated around the same time.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@soGaussian would you be the main consumer of this dataset?
feat(): stores update clients references to new profile clients
This PR aims to update references to firefox_android_clients and firefox_ios_clients to the new new_profile_clients datasets so that the former can be deprecated.