Hi all,
Clients of several of our users suddenly received updates for past meetings, some from April. Microsoft 365 audit logs show the changes came from the Salesforce Inbox / EAC app via Microsoft Graph.
What we found:
- EAC rewrote about 6 months of meetings in batches: first in Salesforce, then in Outlook a few seconds later.
- Only the time zone and the body changed, not times or attendees. Exchange still sent an update to every attendee, since the user is the organizer.
- Trigger identified: IT had emailed users a reminder to [reconnect their account] soon. Users did it, and invites were then resent massively.
- Some users were hit without any mass re-sync. We suspect that any touch on a meeting, for example dismissing a reminder, makes EAC push its version back.
Salesforce Support says this is a known limitation with Microsoft Graph, and notifications can't be suppressed.
Questions:
- Has anyone seen this after an EAC reconnection?
- Is there a way to reconnect without a full re-sync of past events?
- What mitigation worked for you, while keeping email capture active?
Thanks!
1 respuesta
- Yes, this is a known pattern after EAC reconnection. When a user reconnects, EAC re-establishes its sync baseline and re-pushes its stored version of events back out. Because the organiser is the user, Exchange treats each rewrite as a legitimate update and notifies every attendee. The timezone/body-only change rewriting 6 months of meetings is the classic signature. You’re not misreading it.
- Reducing the re-sync window is the key lever. EAC syncs events within a configured time range, and past events outside that window aren’t re-pushed. In your EAC configuration, check the sync timeframe setting for events. If it’s set to sync a wide historical window, narrowing it so EAC only syncs recent and future events dramatically limits what gets rewritten on reconnect. You generally don’t need EAC rewriting April’s meetings anyway. Tightening that window before any mass reconnect is the single most effective prevention. A full reconnect with no past-event sync isn’t really exposed as a clean toggle, but narrowing the event sync window gets you most of the way there.
- The mitigation is to control the reconnection itself, not suppress notifications after the fact.
Narrow the event sync window first (above), then have users reconnect. Far fewer events in scope means far fewer rewrites.