Start Trial
PortalMeridianKeystoneRMM
IT Service & Operations Manual

Infrastructure Map

Service and infrastructure relationships, blast-radius visibility, and the topology context teams need during change and incident work.

Audience: Infrastructure and service ownersFocus: Topology visibility and relationship contextStatus: Public manual

Scope

Infrastructure maps only matter if they improve diagnosis and planning. This page keeps the public-safe operating interpretation and excludes private collection or rendering details.

Where the Feature Lives

The shipped infrastructure-map feature spans three routes:

  1. the relevant workflow
  2. the relevant workflow
  3. the relevant workflow

Together they cover dependency mapping, org health review, and large-screen wallboard monitoring.

Route access is now explicit at the frontend shell: - the relevant workflow requires discovery.view - missing access renders an access-denied state before the page starts graph or health fetches

Using the Infrastructure Map

Open the relevant workflow and select an organization first.

The organization picker now loads accessible organizations in paged batches instead of trusting the first API page, so large multi-org accounts can still select later tenants from the shipped route root.

Available controls

  • Minimum confidence slider
  • Show infrastructure edges toggle
  • Include SNMP devices toggle
  • Host search
  • Device search
  • Zoom and fit controls
  • Edge explorer

What you will see

  • Host nodes
  • SNMP device nodes
  • Edges representing observed or manual relationships
  • Graph stats for nodes, edges, hosts, and devices

If the graph is small enough, the page renders a full Cytoscape graph immediately.

Exploring Large Topologies

When the organization exceeds the graph threshold, the page switches to metadata_only mode.

In that mode: - the graph canvas starts empty - the stats panel still shows total node and edge counts - ranked suggested starting points appear from topology connectivity - you can also begin exploration from host or device search

Metadata exploration workflow

  1. Click a suggested starting point, or search for a host/device if you need a specific entry point.
  2. Load that node and its connected edges into the canvas.
  3. Click the node on the canvas to inspect it.
  4. Use Add Related Nodes to expand one hop outward.

Inspecting a Node

Click any visible node on the canvas to open the right-hand detail panel.

The panel shows: - label - current status - node type - IP address, when present - OS type for hosts - device type for SNMP devices - connected edges

Use Add Related Nodes to pull directly connected nodes and edges into the current canvas.

Large-topology note: - node/device expansion is intentionally bounded per request (default 200 connected edges) - when a node has more connected edges than the current limit, the UI warns that results were truncated

Creating a Service Group from the Map

This workflow is for host selections only.

Steps

  1. Select at least two host nodes on the canvas.
  2. Click Create Service Group.
  3. Review the suggested tier layout.
  4. Rename tiers as needed.
  5. Move hosts between tiers if necessary.
  6. Enter the service-group name and optional description.
  7. Choose whether to enable health tracking.
  8. Create the service group.

What the backend creates: - the service group - one host group per tier - one service-tier row per tier - dependency edges between tiers

SNMP devices can be inspected on the map, but they are not part of the tier-suggestion workflow.

Managing Manual Edges

Use Edges to open the edge explorer drawer.

What you can do there

  • filter edges by discovery source
  • filter edges by protocol
  • page through edge results
  • create a manual edge
  • delete a manual edge

Creating a manual edge

  1. Click Add Edge.
  2. Choose source type and search for the source host/device.
  3. Choose target type and search for the target host/device.
  4. Set protocol and optional target port.
  5. Create the edge.

Manual edges are stored with full confidence and can be removed later. Auto-discovered edges cannot be deleted from this UI. The source/target pickers query the backend as you type, scoped to the selected organization, so you can find nodes beyond the first page of inventory.

Deleting a manual edge

  1. Open Edges.
  2. Find the manual edge row you want to remove.
  3. Click the delete action.
  4. Review the confirmation dialog, which names the source and target endpoints.
  5. Confirm the delete.

Only manual edges can be removed from this workflow. Auto-discovered edges remain under engine control and do not expose a delete action here.

Using Environment Health

Open the relevant workflow to review one organization at a time.

The Environment Health organization picker uses the same paged organization-loading contract as Infrastructure Map, so later accessible orgs are not dropped from the selector in larger accounts.

Page behavior

  • Select an organization
  • Review the composite score
  • Review widgets sorted from worst score to best
  • Click a widget to drill into the related surface
  • Launch War Room from the top-right action

The current widget set covers: - hosts and agents - alerts - patches - backups - certificates - compliance - applications - vulnerabilities - Active Directory - network devices - ransomware - drift - M365

Batch 02 canonicalized the widget drill-down routes so the shipped health page now follows the current router table instead of stale aliases.

Current navigation note: - Environment Health is routed and usable, but it is still a direct URL feature rather than a first-class sidebar destination.

Using War Room

War Room is intended as a wallboard or active watch screen.

Routes: - the relevant workflow

Multi-org mode

Open the relevant workflow with no query string to rotate through all accessible active orgs.

Behavior: - refresh every 30 seconds - rotate to the next org every 15 seconds - highlight newly critical widgets with a flash state - multi-org mode fetches the relevant workflow and rotates from that returned payload instead of loading each org again on every rotation - click any widget card to pivot directly into its canonical drill-down page with org context preserved

Single-org mode

Single-org refreshes also read from the same short-TTL per-org health cache used by summary.

Use the fullscreen button if you want a true wallboard presentation.

Permissions Reference

Permission Operations
discovery.view View map graph, node detail, and edge explorer
discovery.manage Create service groups from the map and manage manual edges
monitoring.view View environment health and War Room data

Current Operational Limits

  • Health summary and War Room use an in-process per-org cache with TTL, not a shared cross-process cache or persisted summary table.

Cross-References

  • docs/architecture/infrastructure-map.md
  • docs/functional/infrastructure-map.md
  • docs/manual/network-discovery.md
  • docs/manual/snmp.md