1184082 - Verify the send logic didnt break submission behavior

Status: RESOLVED WORKSFORME

(Toolkit :: Telemetry, defect, P2)

Description

[Georg Fritzsche [:gfritzsche]]

We need to verify that the send logic changes in bug 1156712 didn't break ping submissions.
I think we mostly need to look at the submission volume pattern over time here.

A monitor like this might answer most of this question, but doesn't actually render data further back:
https://pipeline-prototype-cep.prod.mozaws.net/#sandboxes/PrototypeSandbox-kparlante_ByChannelHour/outputs/PrototypeSandbox-kparlante_ByChannelHour.nightly.cbuf

Comment 1

[Georg Fritzsche [:gfritzsche]]

Katie, can you look into this?

Comment 2

[Georg Fritzsche [:gfritzsche]]

It would also be good to look at the "lag time" development (i.e. average/... time from ping creation to submission) to verify that
* we didn't regress it
* we actually improved it

Comment 3

[Georg Fritzsche [:gfritzsche]]

Mark pointed me to this too:
http://ec2-50-112-66-71.us-west-2.compute.amazonaws.com:4352/#sandboxes/TelemetryChannelMetrics60DaysAggregator/outputs/TelemetryChannelMetrics60DaysAggregator.nightly.cbuf

From a quick first look, this seems sane - note that per mark, the drop after Jul 13 is expected for other reasons.
The first spike is probably catching up on "all the backlogged pings", then things normalize a bit.
The less even distribution is expected from my side (as we submit much more agressively/immediately).

However, i don't think i can judge things properly this week, so i'd love more eyes on this.

Comment 4

[Mark Reid [:mreid]]

(In reply to Georg Fritzsche [:gfritzsche] [away jul 13-17] from comment #3)
> From a quick first look, this seems sane - note that per mark, the drop
> after Jul 13 is expected for other reasons.
See Bug 1172996 for the reason for the drop.

Comment 5

[Katie Parlante]

https://kibana.shared.us-west-2.prod.mozaws.net/#/dashboard/elasticsearch/Telemetry%20Nightly%20By%20Reason

Comment 6

[Thomas Huelbert]

to Katie, moving to [40b9]

Comment 7

[Georg Fritzsche [:gfritzsche]]

Thanks Katie, that is great.
This seems to go fine from a quick look, anyone else notice something suspicious?