There are occasional misfires about data missing by the alert for the "DispatchChanges Normal job backlog time (p50, 15min)" metric. Most recently on Tue, 17 Oct 2023 18:15:38 +0000.
When looking at the actual data for that metric around that time, it looks complete.

Also, there is consistently an email resolving the alert 5 or 10 minutes later.

Notes:

• There is also a task about this alert with respect to the eqiad->codfw datacenter switch: T330770: Investigate DispatchChanges Normal job backlog time (mean avg, 15min) alert post datacenter switch
• This may or may not be similar in cause or intermediate symptoms to T348831: [WD-ORG] [TECH] Max Lag alerts misfire with a DataSource error
• In the past, there have also been genuine issues with data missing: T341054: Wikibase DispatchChanges job potentially broken