Datto RMM Sites

Datto RMM site management: site hierarchy and identifiers, proxy and patch-window settings, site-scoped device/alert queries, and create/update/delete operations for client locations.

SKILL.md

GitHub

Datto RMM Site Management

Overview

Sites in Datto RMM represent client organizations or locations. Each site contains devices, has its own settings, and can have site-level variables. Sites provide organizational hierarchy and enable scoped operations - alerts, jobs, and reports can all be filtered by site.

Anti-triggers

  • The client as a billing entity — use autotask-crm; as a documentation entity, itglue-organizations; as a SOC tenant, rocketcyber-accounts. None of these share IDs with a Datto RMM site.
  • The endpoints themselves rather than the grouping — use datto-rmm-devices.

Key Concepts

Site Hierarchy

Account
└── Sites (many)
    └── Devices (many per site)
        └── Alerts, Jobs, Audit Data

Site Types

Sites can represent:

  • Client companies - External customers
  • Internal locations - Your own offices
  • Projects - Temporary groupings
  • Departments - Internal divisions

Site Identifiers

Identifier Type Description
siteUid string Globally unique identifier
siteId integer Legacy numeric ID
name string Display name

Field Reference

A Site carries identifiers, onDemand/splapiEnabled flags, optional proxySettings, device/alert counts, and a settings object (auto-patch approval, patch window, timezone). See references/fields.md for the complete interfaces.

Common Workflows

Site Lookup by Name

async function findSiteByName(client, name) {
  const response = await client.request('/api/v2/sites?max=250');
  const sites = response.sites || [];

  // Exact match first
  const exact = sites.find(s =>
    s.name.toLowerCase() === name.toLowerCase()
  );
  if (exact) return { found: true, site: exact };

  // Partial match
  const matches = sites.filter(s =>
    s.name.toLowerCase().includes(name.toLowerCase())
  );

  if (matches.length === 0) {
    return { found: false, suggestions: [] };
  }

  if (matches.length === 1) {
    return { found: true, site: matches[0] };
  }

  return {
    found: false,
    ambiguous: true,
    suggestions: matches.map(s => ({
      name: s.name,
      uid: s.uid,
      deviceCount: s.devicesCount
    }))
  };
}

Site health scoring (device status + alert priority weighted into a 0-100 score), a multi-site summary sorted by open alerts, an onboarding checklist validator, and a safe CRUD wrapper are in references/examples.md.

API Patterns

  • GET /api/v2/sites?max=250 - list all sites
  • GET /api/v2/site/{siteUid} - get a single site
  • GET /api/v2/site/{siteUid}/devices?max=250 - devices at a site
  • GET /api/v2/site/{siteUid}/alerts/open / .../alerts/resolved?max=250 - site-scoped alerts
  • POST /api/v2/sites - create a site
  • POST /api/v2/site/{siteUid} - update a site
  • DELETE /api/v2/site/{siteUid} - delete a site (devices become unassigned, not deleted)

See references/api.md for full request/response examples.

Gotchas

  • Deleting a site does not delete its devices - they become unassigned, not removed. Move or reassign devices deliberately before deleting.
  • Site names must be unique - creating a duplicate name fails with 400.
  • Set the site's timezone deliberately - settings.timezone drives when the patch window actually runs.
  • Health scoring is a convention, not an API field - openAlertsCount/devicesCount are raw counts; any "healthy/warning/critical" status is computed client-side (see reference examples).

See references/errors.md for the full site API error table.

Site Naming Conventions

Recommended Format: {ClientName} - {Location/Purpose}

Examples:

  • Acme Corp - Main Office
  • Acme Corp - Remote Workers
  • TechStart Inc - Data Center
  • Internal - IT Department

Related Skills