Sap Landscape Discovery

Environment Configuration

All environment-specific values (subscription ID, AMS workspace ID, proxy URLs, API keys, SAP landscape) are provided via the Team Onboarding instructions. The agent reads these from the onboarding context at runtime. Do not hardcode environment values in this skill.

Authentication: Use the agent's built-in tools (RunAzCliReadCommands, GetArmResourceAsJson, QueryLogAnalyticsByWorkspaceId, GetMetricTimeSeriesElementsForAzureResource) for Azure API calls. These authenticate automatically via the agent's Managed Identity.

Data Reuse (AAU Optimization): Before calling any API or proxy, check if the data was already retrieved earlier in this conversation. Reuse landscape registry, VM power states, config files, and AMS query results from context. Do not re-fetch data that is already available.

Config reads & proxy fallback: Stored SAP/OS configs are read directly from the sap-configs blob container using the agent's own Managed Identity (--auth-mode login) — there is no config proxy. The MCP command proxy is optional and runs only live VM commands; if it is not deployed or errors (timeout, 5xx, unreachable), continue with stored blob configs + Azure-native sources (AMS, ARM API, Azure Monitor). Never block the skill on the proxy.

When to Use

• "What SAP systems do I have?"
• "Show SAP landscape inventory"
• "Show VM power state for all systems"
• "Discover SAP systems in my subscription"
• "Add SAP system AB7 to inventory"
• "Which systems are missing HA monitoring?"

Routing: For full-stack status checks ("Is AB1 up?", "Is SAP running?") → route to SAP Operational Health. Use this skill only for inventory questions, VM power state queries in bulk, or system discovery.

Modes

Mode 1: Auto-discover from Azure

Query Azure Resource Graph for VMs with SAP-related tags and ACSS Virtual Instances.

Mode 2: Import from CSV/conversation

Customer provides data as CSV, pasted table, or verbal description.

Mode 3: Power state check

Live ARM API query for VM instance view (running/stopped/deallocated).

Mode 4: Retrieve existing inventory

Read from agent Knowledge Base (primary), or read the landscape inventory blob directly from the sap-configs container with the agent's own MI (fallback).

Authentication

IMPORTANT — Azure API Access: Do NOT use IMDS tokens (169.254.169.254) or ManagedIdentityCredential — they are not available in the agent sandbox. Instead:

• For Azure Resource Manager queries: Use the built-in GetArmResourceAsJson or RunAzCliReadCommands tools
• For Log Analytics queries: Use the built-in QueryLogAnalyticsByWorkspaceId tool
• For metrics: Use the built-in GetMetricTimeSeriesElementsForAzureResource tool
• For live VM commands: invoke the SAP Command Runner skill (the sap-sre-proxy MCP connector) — never make direct HTTP or proxy calls

# Use the built-in tools above for Azure API access.
import json

# SUB_ID: Use subscription_id from Team Onboarding

# Landscape inventory (Mode 4): read from the agent Knowledge Base (primary), or the
# landscape-inventory blob in the sap-configs container with the agent's own MI (fallback):
#   az storage blob download --auth-mode login --account-name <sa> --container-name sap-configs \
#     --name landscape/sap-landscape-inventory.json
# There is NO /api/registry endpoint — the proxy is MCP (live commands only).

Mode 1: Auto-discover

# Pseudocode — use GetArmResourceAsJson or RunAzCliReadCommands tool instead
def discover_from_tags():
    query = """
    resources
    | where type == 'microsoft.compute/virtualmachines'
    | where tags contains 'SAPSystemSID' or tags contains 'sapsid' or tags contains 'SID'
    | project name, resourceGroup, location, tags,
              zones=zones, vmSize=properties.hardwareProfile.vmSize
    """
    return query_resource_graph(query)

# Pseudocode — use GetArmResourceAsJson or RunAzCliReadCommands tool instead
def discover_from_vis():
    try:
        vis_list = arm_get(f"/subscriptions/{SUB_ID}/providers/Microsoft.Workloads/sapVirtualInstances", "2023-10-01-preview")
        return vis_list.get("value", [])
    except Exception:
        return []

Mode 3: Power State Check

# Pseudocode — use GetArmResourceAsJson or RunAzCliReadCommands tool instead
def check_power_state(vm_name, rg):
    iv = arm_get(f"/subscriptions/{SUB_ID}/resourceGroups/{rg}/providers/Microsoft.Compute/virtualMachines/{vm_name}/instanceView")
    statuses = iv.get("statuses", [])
    for s in statuses:
        if s.get("code", "").startswith("PowerState/"):
            return s["code"].replace("PowerState/", "")
    return "unknown"

Topology inference (populate architecture + deployment)

When building or refreshing the inventory, classify each system into the canonical topology so downstream skills adapt (see the _field_reference in sap-landscape-inventory.json). Infer from discovered resources:

• architecture — count active HANA DB nodes for the SID: one DB node per replica → scale-up; a master + worker set (multiple DB VMs sharing one HANA SID/instance, HANA scale_out role tags, or >1 DB VM in the same HSR site) → scale-out (record scale_out.db_worker_nodes / standby_nodes).
• deployment — decide in this order:
  1. A DB replica (or app tier) in a second Azure region with async HSR / ASR → disaster-recovery.
  2. Else a Pacemaker cluster present (2+ DB VMs with HSR, ASCS+ERS pair, internal Standard LB, SBD/fence-agent) → high-availability (set ha.*, mark VMs ha_role: primary|secondary).
  3. Else DB / ASCS / App on separate VMs, no cluster → distributed.
  4. Else all tiers on one VM → standalone.
• Set type = "<architecture>-<deployment>" (e.g. scaleout-ha). Populate each VM's roles, ha_role, and (scale-out) node_role.
• If you cannot confidently classify (e.g. missing tags), record your best guess and flag it in the report so the user can correct it in the inventory file.

Output Format

Display inventory as a structured table: | SID | Topology (arch-deployment) | RG | Region(s) | VMs | Roles | VM Size | Zones | DB nodes (master/worker/standby) |

For power state: append a column with GREEN (running), RED (stopped/deallocated).