Skip to content
bzonfhir docs

03 — Channel Detail (statistics + recent messages)

Mirth Connect dashboard → expand a channel row, OR Channels → double-click a deployed channel. The drill-down view for a single channel.

What this screen shows in Mirth

When you expand a channel row on the Dashboard (clicking the ”+” icon to its left), an inline panel slides open underneath the row revealing per-channel statistics broken out by Source and each Destination:

  Channel: HL7-ADT-to-FHIR-Patient
  ├── Source (LLP Listener)        Received: 12,340   Filtered: 12       Queued:   0   Errored: 3
  ├── Destination 1: HTTP Sender   Sent: 12,300       Errored: 25                       Queued: 12
  └── Destination 2: File Writer   Sent: 12,328       Errored: 0                        Queued: 0

Each row shows the same counter set as the Dashboard, but scoped to that connector. There’s a small Show Messages button per connector that opens the Message Browser pre-filtered to that destination.

If you instead open the channel by double-click, you land on a channel-detail window with three tabs:

  1. Summary — small set of fields: name, description, deploy state, source connector type, destination count.
  2. Statistics — full historical message counters with reset buttons.
  3. Messages — embedded Message Browser pre-filtered to this channel (this is the actual usable tab).

A footer bar shows: last received timestamp, last sent timestamp, queue depth across all destinations, “Connection State” per connector (Idle / Connected / Polling / Disconnected with a reason string).

What we’re adapting

The per-Interface page (/interfaces/[id]) is the right home for:

  • A header card with: name, source system (e.g. “Epic ADT feed”), state badge, last activity timestamp, total queue depth
  • A breakdown by destination/subscription — each subscription gets its own row with: Sent, Errored, Pending, last delivered timestamp, link to this subscription’s history (which is #404’s screen)
  • A “Recent Messages” embedded list — last ~50 messages this Interface has seen, time-ordered desc, with a “View all → message browser pre-filtered” link
  • Source-side metrics: receive rate (msg/min over last 5/15/60 min), inbound connection state, last received time
  • A small set of operator actions: Pause, Resume, Drain queue, Replay last N

The pattern of “open this screen and within 5 seconds know if this interface is healthy” is exactly what Mirth’s drill-down gets right.

What we’re explicitly NOT copying

  • The triple-tab layout (Summary / Statistics / Messages). We collapse it to one page with sections.
  • The “Reset Statistics” buttons. Our counters live in Prometheus, scoped to windows — no manual reset.
  • The inline “expand row” interaction on the dashboard. We use a navigation click. Expanding rows is fiddly and doesn’t scale to deep links.
  • The “Channel Editor” launched from this view. Out of scope for v1.
  • Per-destination “Reprocess Queued” button — replay actions live in the DLQ screen (#403), not here.

Which ticket implements our equivalent

#401 (UI: per-interface state and stats). Secondary readers: #400 for the “embedded recent-messages” pattern, and #404 for the per-subscription delivery-log linked from this page.