Academy/VCP-VCF 9.0 Administrator (2V0-17.25)/Lab: Configure VCF Operations with NSX Adapter and Custom Dashboards
This lab targets VCF 9.0

Lab: Configure VCF Operations with NSX Adapter and Custom Dashboards

VCF 9.0Intermediateadmincloud-opsnetwork⏱ 75 min

Objectives

  • Add NSX adapter to VCF Operations for network topology monitoring
  • Create custom dashboard with NSX-aware widgets (DFW analytics, segment health)
  • Configure cross-component correlation between vCenter, NSX, and vSAN metrics
  • Differentiate metrics, properties, and logs with hands-on examples

Prerequisites

VCF management domain with VCF Operations deployed and NSX Manager operational

Prior labs: vcp-admin-01

Required skills:

  • VCF Operations UI
  • NSX Manager basics

Lab Environment

Holodeck VCF pod with management domain, NSX Manager cluster, VCF Operations.

Credentials

SystemUsernamePassword
VCF Operationsadmin

Tasks

Task 1 Add the NSX Account and Verify Network Metrics Collection

manageability

NSX adapter enables network topology visualization and DFW analytics — key VCF-only capability not in VVF.

Step 1

In VCF Operations go to Administration > Integrations, open the Accounts tab, click Add and choose NSX. Enter a Name, the NSX Manager VIP/FQDN and a credential: either NSX Credentials (a dedicated read-only user such as one with the Auditor role, which collects all data except the load balancer) or NSX Client Certificate Credentials (Principal Identity). Select the collector group, click Validate Connection, accept the certificate if prompted, then Add. In a VCF instance the NSX account may already exist from the instance connection — review it rather than adding a duplicate.

NSX account appears on the Accounts tab with a green collection status. Collection begins on the next cycle.
Roles and what they collect (VCF 9.0.1 docs p.3111): Enterprise Admin all data; Security Engineer / Security Operator / Auditor all data except the load balancer; Network Engineer no load balancer and limited router data; LB Admin / LB Auditor / Netxpartner Admin cannot collect. Use a dedicated account, not the admin login.
Step 2

Wait a few collection cycles (5-minute cycles; allow 10-15 minutes). Open the NSX adapter instance object (Environment > Object Browser, or the account's object link) and verify Tier-0/Tier-1 gateways, segments, Edge nodes and transport nodes appear with metrics beginning to populate.

NSX inventory shows Tier-0 gateways, Tier-1 gateways, Segments, Edge nodes and Transport nodes with metrics populating.
Initial collection takes 10-15 minutes for a small environment. Large NSX deployments (1000+ segments) may take 30+ minutes.
Step 3

Confirm the NSX management pack dashboards are available: Visualize > Dashboards, filter on 'NSX'. The management pack ships NSX Configmax Metrics, NSX Inventory, NSX Edge, NSX Segment Inventory, NSX Transport Node Performance, NSX Segment Performance and NSX Load Balancer Performance dashboards (VCF 9.0.1 docs, 'Dashboards in NSX Management Pack'). The pack is part of VCF Operations in 9.0; no separate content pack install is needed ('Content Pack' is VCF Operations for logs terminology).

NSX dashboards listed: NSX Configmax Metrics, NSX Inventory, NSX Edge, NSX Segment Inventory, NSX Transport Node Performance, NSX Segment Performance, NSX Load Balancer Performance.
If the dashboards are missing, the NSX account has not collected yet or the user's dashboard sharing excludes them; check Administration > Integrations status first.
Step 4

Open the NSX Inventory and NSX Edge dashboards. Review the widgets: object counts (gateways, segments, transport nodes), Edge node health and resource utilisation. Then open NSX Configmax Metrics and compare current object counts against the configuration maximums.

NSX Inventory shows counts of gateways, segments and transport nodes; NSX Edge shows per-Edge health and CPU/memory; NSX Configmax Metrics shows utilisation against limits.
Configmax tracking is the design-relevant widget: it tells you how close the NSX deployment is to a scale ceiling before it becomes an outage.

Validation Gate

Check: Administration > Integrations > Accounts shows the NSX account collecting; the NSX adapter inventory shows Tier-0 gateway, segments and Edge nodes; the NSX management pack dashboards open with data

Expected: NSX inventory populated with all managed objects and the NSX dashboards show current values

Common Errors

Using admin credentials instead of service account for NSX adapter — password rotation breaks collection
Using the NSX admin account for collection instead of a dedicated read-only (Auditor) account or a Principal Identity certificate
Expecting immediate data after adapter addition — initial collection takes 10-15 minutes

Task 2 Create Cross-Component Dashboard

manageability

Cross-component dashboards demonstrate the VCF Operations value proposition — correlating compute, network, and storage metrics in a single view.

Step 1

Create new dashboard: 'VCF Infrastructure Health'. Add widgets: (1) Heatmap — hosts by CPU, (2) NSX segment traffic chart, (3) vSAN latency trend, (4) Top-N VMs by network throughput.

Dashboard layout with 4 widgets: CPU heatmap (green-to-red color scale), NSX segment traffic line chart, vSAN latency trend line, and Top-N VM table.
Design dashboards for specific personas: executive (high-level health), operator (detailed metrics), security (DFW analytics).
Step 2

Configure widget interactions: Host heatmap → filters NSX segment widget to show segments on selected host. Top-N → links to VM detail.

Widget interactions configured: clicking a host in heatmap filters the NSX segment widget to show only segments with VMs on that host.
Widget interactions make dashboards actionable. Without interactions, each widget is an isolated view that requires mental correlation.
Step 3

Add alert list widget filtered to Critical/Immediate alerts across all adapters (vCenter + NSX + vSAN).

Alert list widget shows Critical and Immediate alerts from all configured adapters. Alerts include source (vCenter/NSX/vSAN), object name, and triggered time.
Filter alert widgets carefully — showing all alerts creates noise. Focus on Critical and Immediate for operator dashboards; include Warning for capacity planning dashboards.
Step 4

Save and share dashboard with operator role (read-only access).

Dashboard shared with operator role. Operators can view but not modify the dashboard layout or widget configurations.
Dashboard sharing supports role-based access. Create different dashboards for different roles rather than one dashboard for everyone.

Validation Gate

Check: Dashboard 'VCF Infrastructure Health' loads with all widgets showing data from vCenter, NSX, and vSAN

Expected: Cross-component widgets populated, interactions working

Common Errors

Creating dashboards with too many widgets — cognitive overload reduces effectiveness; limit to 6-8 widgets per dashboard
Not configuring widget interactions — isolated widgets require manual correlation across separate views
Sharing dashboards with edit access to operators — accidental modification loses carefully designed layouts

Task 3 Demonstrate Metrics vs Properties vs Logs

manageability

Understanding the three data types is an explicit exam objective. Hands-on differentiation builds exam confidence.

Step 1

Select any ESXi host → Metrics tab. Identify: CPU Usage % (metric — numeric, changes every 5 min). Note collection interval and retention.

Metric example: CPU Usage % = 45.2 (numeric, sampled every 5 minutes, stored for 6 months default retention).
Metrics drive alerts and capacity planning. Know the collection interval (default 5 min) and retention period for sizing VCF Operations storage.
Step 2

Same host → Properties tab. Identify: ESXi Version (property — static, changes only on upgrade). Note: properties enable filtering and grouping.

Property example: ESXi Version = 9.0.0 (string, updated only when host is upgraded, used for filtering and grouping in dashboards).
Properties enable powerful grouping: filter all hosts by ESXi version to identify upgrade candidates, or group by hardware model for lifecycle planning.
Step 3

Navigate to Operations for Logs → Explore Logs → filter by source = same ESXi host. Identify: Log entry (log — text event, timestamped, used for troubleshooting).

Log example: 2025-01-15T10:23:45Z esxi-01 vpxa: [info] Task completed successfully (text event with timestamp, source, and message body).
Logs are the troubleshooting backbone. VCF Operations for Logs provides structured search and log-based alerting — essential for proactive incident detection.
Step 4

Document the three data types with one example each. This is a common exam question format.

Documentation showing three data types with examples from the lab. This format matches common exam question patterns.
Exam tip: metrics are numeric and time-series, properties are descriptive and semi-static, logs are text events used for troubleshooting. Remember: metrics for alerting, properties for inventory, logs for root cause.

Validation Gate

Check: Can articulate metric vs property vs log with concrete examples from the lab environment

Expected: Clear understanding demonstrated through hands-on exploration

Common Errors

Confusing metrics with properties — metrics change frequently (CPU usage), properties change rarely (ESXi version)
Not understanding log retention vs metric retention — different retention policies for different data types
Ignoring the relationship between data types — a metric alert triggers investigation in logs, properties filter the scope

Task 4 Configure Alert Definitions and Notification Channels

availability

Alert configuration transforms VCF Operations from a passive monitoring tool to an active operational assistant. Proper alerting reduces MTTR and prevents incidents.

Step 1

From the left menu click Infrastructure Operations > Configurations, then the Alert Definitions tile (VCF 9.0.1 docs p.3224). Review existing alerts for NSX objects. Key alerts: NSX Manager cluster health degraded, Edge node CPU >80%, DFW rule push failure, Transport node disconnected. Enable notification for each critical alert.

Alert definitions list shows NSX-specific alerts from the Content Pack. Each alert has: name, criticality (Critical/Immediate/Warning/Info), trigger condition, and recommendation.
Focus on high-impact alerts first: NSX Manager cluster health and Edge node availability directly affect network data plane.
Step 2

Create a custom alert: 'vSAN Capacity Warning'. Configure: Metric=vSAN Used Capacity %, Condition=> 70%, Criticality=Warning, Wait Cycles=3 (confirms trend over 15 min). Add recommendation text: 'Review capacity trending and plan expansion. Check for orphaned snapshots consuming space.'

Custom alert created and visible in Alert Definitions. After 3 collection cycles above threshold, alert triggers and appears in the Alerts dashboard.
Wait Cycles prevent false alarms from transient spikes. 3 cycles = 15 minutes at 5-min collection interval. Increase for less critical alerts.
Step 3

Configure notifications in two parts. (1) Outbound instances: from Infrastructure Operations > Configurations, click the Outbound Settings tile (VCF 9.0.1 docs p.3206) and add outbound plug-in instances: Standard Email Plugin with your SMTP server, Webhook Notification Plugin for ServiceNow/Jira (or the Service-Now Notification Plugin), and optionally SNMP Trap or Slack. Test each instance before saving. (2) Notification rule: from Infrastructure Operations > Configurations, click the Notifications tile (p.3197), click Add and create a rule that sends Critical and Immediate alerts to ops-team@company.com through the email outbound instance. Documented plug-in types are Log File, Rest Notification (deprecated), Standard Email, SNMP Trap, Webhook Notification, Slack and Service-Now Notification (VCF 9.0.1 docs p.3199, p.3202, p.3205).

Outbound instances configured and tested; notification rule in place. Test notification email received by operations team. Webhook integration documented for future ServiceNow/Jira integration.
Filter notifications aggressively — sending all alerts causes alert fatigue. Ops team email gets Critical/Immediate only. Warning alerts go to a dashboard for weekly review.
Step 4

Design an alert escalation matrix: Tier 1 (Info/Warning) — dashboard review during daily standup. Tier 2 (Immediate) — email to on-call engineer, 30-minute response. Tier 3 (Critical) — email + webhook to PagerDuty/ServiceNow, 15-minute response. Document the matrix and map VCF Operations alert levels to organizational response tiers.

Alert escalation matrix with 3 tiers mapping VCF Operations criticality to organizational response procedures. Each tier has: notification method, response time, escalation path.
VCDX design: alert escalation must map to SLA commitments. If the SLA requires 15-minute response, the alert must reach the engineer within 5 minutes to allow 10 minutes for diagnosis.

Validation Gate

Check: Custom alert created, notification channel configured, escalation matrix documented

Expected: Alert fires on threshold breach, notification reaches operations team, escalation matrix ready for production use

Common Errors

Not setting Wait Cycles on alerts — single-sample alerts trigger on transient spikes causing alert fatigue
Sending all alert levels to email — operators ignore alerts when volume is too high
Not testing notification channels — alerts fire but nobody receives them due to SMTP misconfiguration

Task 5 Create Super Metrics for Cross-Component Analysis

manageability

Super Metrics combine data from multiple objects into custom calculated metrics — enabling business-relevant KPIs that no single adapter provides.

Step 1

From the left menu click Infrastructure Operations > Configuration, then the Super Metrics tile (VCF 9.0.1 docs p.4178). Create a new Super Metric: 'Cluster CPU Overcommit Ratio'. Formula: Sum of VM vCPU across all VMs in cluster / Total physical cores in cluster. Assign to cluster object type.

Super Metric calculates in real-time: for a 4-host cluster with 128 physical cores and 200 total vCPU assigned, ratio = 1.56:1. This overcommit ratio is a key capacity planning metric.
CPU overcommit ratios vary by workload: 3:1 acceptable for VDI, 1.5:1 for production servers, 1:1 for latency-sensitive applications. Document the target ratio in your VCDX design.
Step 2

Create a second Super Metric: 'vSAN Days to 70%'. Formula: (0.70 x Total vSAN capacity - Current used) / Average daily growth (30-day trend). Assign to the vSAN cluster object type. Compare the result with the cluster's native Capacity > Time Remaining value, which is the documented capacity-engine answer to the same question.

Super Metric shows estimated days until vSAN reaches the 70% capacity threshold. Example: 10 TB total, 5 TB used, 10 GB/day growth: (7 TB - 5 TB) / 10 GB per day = 200 days to 70%.
This metric drives procurement planning. Alert when days remaining < 90 to allow time for hardware ordering and installation.
Step 3

Add both Super Metrics to the VCF Infrastructure Health dashboard created in Task 2. Create widgets: Scoreboard showing overcommit ratio (green <2:1, yellow 2-3:1, red >3:1) and Trend chart showing vSAN days to 70% over last 30 days.

Dashboard now shows business-relevant metrics alongside infrastructure metrics. Executive stakeholders can understand capacity posture without interpreting raw metrics.
Super Metrics bridge the gap between technical infrastructure metrics and business-relevant KPIs. Design them for your audience: executives want trends and forecasts, operators want thresholds and alerts.
Step 4

Document the Super Metric design pattern for VCDX: (a) identify the business question (When will we run out of capacity?), (b) identify the source metrics (vSAN used, vSAN total, daily growth), (c) define the formula, (d) assign to object type, (e) create dashboard widget, (f) configure alert threshold. This pattern can be applied to any cross-component KPI.

Super Metric design pattern documented with 6 steps. This becomes a reusable framework for creating new business-relevant metrics.
VCDX defense: demonstrate that your monitoring design goes beyond default metrics. Custom Super Metrics show operational maturity and business alignment.

Validation Gate

Check: Super Metrics created, dashboard widgets configured, design pattern documented

Expected: Overcommit ratio and vSAN days to 70% visible on dashboard with color-coded thresholds

Common Errors

Creating Super Metrics without defining the business question first — results in metrics nobody uses
Not assigning Super Metrics to the correct object type — metric calculates but does not appear on expected objects
Setting alert thresholds on Super Metrics without Wait Cycles — transient calculation spikes cause false alarms

Final Validation

VCF Operations fully configured with NSX adapter, cross-component dashboards, alerting, and custom Super Metrics.

✓ NSX adapter collecting data → NSX objects visible in inventory with metrics populating

✓ Cross-component dashboard operational → Widgets from vCenter, NSX, and vSAN with interactions

✓ Metrics/Properties/Logs differentiated → Can articulate differences with concrete examples

✓ Alert configuration complete → Custom alerts, notifications, and escalation matrix documented

✓ Super Metrics providing business KPIs → Overcommit ratio and capacity forecast on executive dashboard

Cleanup / Restore

• Take snapshot 'post-vcf-ops-nsx-dashboard'

• Export dashboard definitions for backup

• Save alert escalation matrix document

• Document Super Metric formulas for knowledge transfer

Design Reflection (VCDX)

Panelist: How does VCF Operations with NSX adapter change your monitoring strategy compared to VVF? What network metrics would you alert on for a multi-tier application?

Requirements

  • Unified monitoring across compute, network, and storage
  • Proactive alerting with <15 minute response for critical issues
  • Business-relevant capacity forecasting for procurement planning

Constraints

  • NSX adapter adds additional object count to VCF Operations sizing
  • Super Metric calculation increases VCF Operations CPU overhead
  • Notification channels require outbound network access (SMTP, webhook)

Assumptions

  • NSX Manager accessible from VCF Operations network
  • SMTP server available for email notifications
  • Operations team trained on alert response procedures

Risks

  • Oversized VCF Operations may be needed for large NSX deployments (10,000+ segments)
  • Alert fatigue from poorly tuned thresholds reduces response effectiveness
  • Super Metric formulas may produce misleading results with incorrect source metrics

Self-Assessment Discussion Prompts

  1. How would you size VCF Operations for 10,000 NSX segments?
  2. What DFW metrics indicate a security policy misconfiguration?
  3. How do you design dashboards for different stakeholder audiences?
  4. What Super Metrics would you create for a VCDX design review?

Extensions

What-If and Capacity Planning in VCF Operations

VCF Operations (formerly VMware Aria Operations) includes what-if analysis for workload and infrastructure planning (VCF 9.0.1 docs p.5262). Model adding 50 VMs to the management cluster and removing one host, compare the result with the vSAN Days to 70% super metric, and record which gives the more defensible procurement date.

Custom Content Pack Development

Create a custom Content Pack with dashboards tailored to your organization. Include: executive health summary, operations detail view, security compliance dashboard, and capacity planning view. Export and version-control the Content Pack.

VCF Operations API Automation

Use the VCF Operations REST API to automate: dashboard creation, alert configuration, Super Metric definition, and report generation. Build a Python script that deploys a standardized monitoring configuration to new VCF environments.

⚠ Known Pitfalls

Using admin credentials for NSX adapter instead of a dedicated service account — password rotation breaks metric collection silently
Creating dashboards with 15+ widgets — cognitive overload reduces operational effectiveness; limit to 6-8 focused widgets per dashboard
Not configuring Wait Cycles on alerts — single-sample triggers cause alert fatigue and erode trust in the monitoring system
Ignoring Super Metric opportunities — default metrics answer infrastructure questions but miss business-relevant KPIs like capacity forecast and overcommit ratios

References

Was this page useful?
Type to search. ↑ ↓ to move, Enter to open, Esc to close.