Can we kill the spocs fill event
Categories
(Firefox :: New Tab Page, task, P2)
Tracking
()
People
(Reporter: thecount, Unassigned)
Details
I think we can kill the spocs fill event ping.
We initially did this before launching Discovery Stream, which at the time we were hunting down oddities in our spoc fill rate.
I suspect, if there was a bug with the old AS, Discovery Stream fixed it.
Do we see value in keeping the ping around?
It adds significant complexity and friction to the code around spocs, so there is a benefit on my end for it's removal.
Keeping it around, might have the benefit of catching future unexpected fill issues, but I'm not even fully confident it'll do that very well. The nature of a spocs fill event, has to be built in a way to know what it's going to report, so it'll only catch the known errors, but the likely failure case for this is going to be unknown errors, which spocs fill wouldn't be built to handle. I believe.
Thoughts?
|
||
Comment 1•5 years ago
|
||
/cc :klukas.
Once this ping is discontinued in Firefox, we should also clean it up in the data pipeline.
|
||
Comment 2•5 years ago
|
||
Can we get rid of this now that we have fixed the bug?
|
|
||
Updated•5 years ago
|
|
||
Comment 3•5 years ago
|
||
Agreed that we can kill it! The work for removing it from the client/data pipeline is up to Scott and Nan, though.
|
|
||
Updated•5 years ago
|
|
||
Updated•3 years ago
|
Description
•