Quick start guide for Site Reliability Engineers

Prerequisites

Before you start, confirm the following prerequisites:

• Your IT admin installed Instana agents on the hosts and clusters that you want to monitor.
• You have a user account in Instana with at least User role permissions.
• You have your Instana tenant URL. It looks like https://your-company.instana.io.

Log in and understand the UI

In a browser, open your Instana tenant URL and log in with your credentials. The Home page opens and shows a summary of your system status.

The following areas are most relevant to your work as an SRE:

• Events > Incidents: Active problems that require attention.
• Applications > Services: All discovered services with their error rates and latency.
• Service levels: Service level objectives (SLOs) and error budgets.
• Synthetic monitoring: Proactive checks that verify service availability.
• Infrastructure: Platform and managed services status.

Create application perspectives

Application perspectives group the related services into a single logical application so you can monitor them together instead of one dashboard at a time.

To create an application perspective:

1. Go to Applications > Applications tab.
2. Click + Add to create a new application perspective.
3. Select model - Choose the type of application to model:
   • Services or endpoints - For well-known services that won't change frequently
   • A critical user journey - For monitoring specific user interactions and business transactions
   • Environment or region - For grouping by cloud, zone, or host information
   • An important customer or tenant - For customer-specific dashboards using HTTP parameters or custom tags
   • Kubernetes or container - For grouping by namespace, container name, or platform labels
   • Request attributes - For filtering by HTTP headers, query parameters, or RPC methods
   • Technology - For grouping by database type, Java application, or scripting application
   • Custom tags - For custom metadata added via SDK, platform, or agent
4. Click Next to proceed.
5. Specify application - Define the application scope:
   • Add filters by using tags (for example, database or technology).
   • Select downstream services to include:
     • No downstream services - Source perspective only
     • Immediate downstream database and messaging services - Includes direct dependencies
     • All downstream services - End-to-end view
   • View matched services in real-time.
   • Click Next to proceed.
6. Provide details - Configure the perspective:
   • Application perspective name - Enter a descriptive name
   • Select default dashboard view - Choose between Inbound calls or All calls
7. Click Create. The application perspective appears in the Applications menu. Click it to see aggregated metrics for all included services.

For more information, see Application perspectives.

Set up service level objectives

SLOs define reliability targets for your services and help you track whether you are meeting user expectations and how much error budget remains.

To create an SLO:

1. Go to Service levels and click Create service level objective.
2. Select the entity type: Application perspective, Infrastructure, Synthetic tests, or Website.
3. Choose the specific application, infrastructure type, synthetic test, or website to monitor.
4. Define the scope by selecting filters (optional) to narrow down the monitored entities.
5. Set the indicator (metric to measure), objective (target percentage), and time window.
6. Enter SLO details such as name and description, then click Create.

View your SLOs in the Service level objectives table, which displays entity type, entity name, blueprint type, remaining error budget, and status against target.

Configure synthetic monitoring

Synthetic monitoring runs scripted checks at regular intervals from external locations to verify that your services are available and responding correctly.

To configure synthetic monitoring:

1. Go to Synthetic monitoring.
2. Click Create a synthetic test.
3. Select a test type:
   • API simple: Check availability of a single REST API endpoint.
   • API script: Test multiple REST API calls in sequence.
   • Browser simple: Check a single webpage for HTTP errors.
   • Browser script: Simulate user interactions with your website.
   • DNS: Verify domain name resolution.
   • SSL certificate: Check the certificate expiration.
4. Configure the test and select test locations.
5. Click Create to create the test.

After you create the test, it runs automatically and alerts you if it detects a problem.

For more information, see Synthetic monitoring.

Set up incident alerts

Alert channels connect Instana events to your team's existing communication tools so that on-call engineers are notified when an incident occurs.

To set up an incident alert:

1. Go to Settings > Events & alerts > Alerts.
2. Click New alert.
3. Enter a name for the alert.
4. From the Events drop-down list, select whether to alert on event types, selected events, or application Smart Alerts.
5. Click Add events to add the events that you want to monitor.
6. From the Scope drop-down list, select where the alert applies.
7. Click Add alert channels to add one or more alert channels.
8. Optional: Include the entity name in the alert title.
9. Optional: Add custom payloads.
10. Click Create.

For more information, see Configuring and managing alerts.

Create a team dashboard

Dashboards provide a single view of the metrics that matter most, such as SLO status, active incidents, and service error rates.

To create a team dashboard:

1. Go to Custom dashboards.
2. Click Create dashboard.
3. Enter a name (for example, SRE On-Call Dashboard).
4. Click Add widget and select a widget type (Apdex chart, Big number, Chart: pie, time series, Histogram, Markdown, SLO, SLO Legacy, Table, Time zone, Top list).
5. Click Next and configure the widget by selecting data source, metrics, and display options.
6. Click Create to add the widget to your dashboard.
7. Customize time range, add filters, and arrange widgets as needed.
8. Click Save changes.
9. Share the dashboard with your team by adjusting permissions.

For more information, see Building custom dashboards.

Investigate incidents

When an incident occurs, use Instana to identify the root cause and affected services. Traces, logs, and infrastructure metrics are in one place and automatically linked.

To investigate an incident:

1. Go to Events > Incidents.
2. Click the incident to open the incident details page.
3. Review the following information:
   • Triggering event: The event description and triggering entity that initiated the incident.
   • Affected entities: The number and list of entities that the incident impacts.
   • Metric violations: Status graphs and metrics that exceeded thresholds.
   • Related events: More events associated with the incident.
4. Click Analytics to drill deeper into traces and logs.
5. Go to Analytics > Applications > Calls and filter by the affected service and time range to find specific failing requests.
6. Click a failing request to see the full trace with timing for each step.
7. Use the trace to identify which service or database query is causing the problem.

For more information, see Analyzing incidents.

Set up automation actions

Automation actions run automatically when specific events occur, reducing manual effort and response time. You can create user-defined actions or use AI-generated actions.

To set up automated actions, see Using the automation framework in Instana.

Analyze platform health

Monitor the health of your platform infrastructure to make sure that applications have the resources they need.

To analyze platform health:

1. Go to Infrastructure > Map to see a visual representation of your infrastructure.
2. Click a host or container to see detailed metrics:
   • CPU usage
   • Memory usage
   • Disk I/O
   • Network traffic
3. Go to Infrastructure > Comparison Table to compare metrics across multiple hosts side by side.
4. Look for resource bottlenecks that might affect application performance.

For more information, see Analyze Infrastructure.

Review and optimize

Regularly review your monitoring setup to ensure that it remains that effective.

To optimize your monitoring, complete the following tasks:

• Weekly: Review open incidents and identify patterns that indicate recurring problems.
• Monthly: Review SLO and adjust targets or improve services as needed.
• Quarterly: Review alert effectiveness and reduce alert fatigue by tuning thresholds or consolidating alerts.

Next steps

After you complete the initial setup, consider the following next steps:

• Expand synthetic monitoring: Add more synthetic tests to cover critical user journeys and API endpoints. For more details, see Synthetic monitoring.
• Refine SLOs: Work with product teams to define SLOs that align with user expectations. For more details, see Service level objectives.
• Build runbook automation: Identify recurring incidents and create automated responses to reduce manual intervention. For more details, see Automation.
• Integrate with incident management: Connect Instana to your incident management platform (PagerDuty, ServiceNow, or Jira) for seamless incident workflows. For more details, see ServiceNow alert channel.

Quick reference: Key locations for SRE tasks