Hi all,
After the Workflows app updated to 1.3537.6 (published 4 Nov 2026), our custom Davis problem → ServiceNow workflow stopped working. The cause was a change in the trigger event that I can’t find in the release notes, which only mention expression autocomplete previews and the new maintenance-window option for Davis problem triggers.
What I see:
- Before: event()[“event.id”] was the internal problem ID, e.g. -7228652111118287103_1794343200325V2
- Now: event()[“event.id”] is the display ID, e.g. P-2611101
- The internal ID is now only available as classic_problem_id (dt.davis.problems / trigger payload), and dt.davis.event_ids holds the underlying Davis event ID(s)
Side effects:
- problemsClient.getProblem({ problemId: event.id }) fails, because the classic API doesn’t accept P-2611101.
- The classic API’s display ID for the same problem is different (P-26111 vs P-2611101 in Grail), so a displayId(“…”) selector lookup returns nothing.
- CUSTOM_ANNOTATION events with annotation.problem_ids = no longer show on the problem in the new Problems app. They only appear when annotation.problem_ids = the P-… display ID.
- Problem links built as /ui/apps/dynatrace.davis.problems/problem/ now show “Failed to load problem”. /problem/P-… works.
I’ve worked around it. But we’d like to confirm:
- Is the event.id change intentional, and is it going to stay?
- Which identifier should workflows use as a stable, long-term problem ID for external systems: event.id, display_id or classic_problem_id?
- Is classic_problem_id guaranteed to remain in dt.davis.problems and the trigger payload?
- What is the supported way to attach annotations/comments to a problem in the new Problems app from a workflow?
- Why do the classic and Grail display IDs differ for the same problem?
- Could the release notes be updated to call this out? Other custom workflows using getProblem(event.id) will break the same way.
Thanks,
Sujit