Setting up the first SLO

Vsceptre Team

May 10, 2023

6 min read

NewsObservability

This is the final piece of the 3 part series "The path to your first SLO". We have discussed on the basics of what to observe and how to get the relevant metrics in part 1 and part 2 of this series. This time we are going to have a quick look on to setup a simple service availability monitoring SLO with Nobl9and SolarWinds Pingdom.

Series Overview: The Path to Your First SLO

This comprehensive 3-part series guides you through building an effective observability practice, from identifying the right metrics to implementing your first Service Level Objective.

Part 1: How to identify the golden metrics for SRE

Part 2: How to obtain the metrics for SLO tracking

Part 3: Setting up the first SLO (Current)

Prerequisites

Before setting up your first SLO, ensure you have:

Identified your golden metrics (from Part 1)
Established data collection methods (from Part 2)
Access to Nobl9 platform
SolarWinds Pingdom account for monitoring

SLO Setup Process

Step 1: Define Service Availability

Start by defining what "availability" means for your service. This typically includes:

HTTP response codes (200-299 range)
Response time thresholds
Error rate limits

Step 2: Configure Nobl9

Set up your SLO in Nobl9 with the following configuration:

# Nobl9 SLO Configuration

service: "web-application"

slo_name: "availability"

objective: 99.9%

time_window: "30d"

data_source: "pingdom"

Step 3: Integrate SolarWinds Pingdom

Connect Pingdom as your data source for monitoring:

Configure Pingdom checks for your endpoints
Set up appropriate check intervals
Configure alerting thresholds
Export metrics to Nobl9

Best Practices

Do's

• Start with simple, measurable metrics
• Set realistic objectives (99.9% vs 100%)
• Monitor SLO burn rate regularly
• Document your SLO decisions
• Review and adjust objectives quarterly

Don'ts

• Don't set unrealistic objectives
• Avoid too many SLOs initially
• Don't ignore SLO violations
• Avoid setting SLOs without stakeholder buy-in
• Don't forget to monitor error budgets

Monitoring and Alerting

Once your SLO is set up, it's crucial to monitor its performance and set up appropriate alerting:

Alerting Strategy

Warning Level: 80% of error budget consumed
Critical Level: 95% of error budget consumed
Emergency Level: Error budget exhausted

Next Steps

Congratulations! You've successfully set up your first SLO. This is just the beginning of your observability journey. Consider these next steps:

• Expand SLO coverage to more services
• Implement SLIs (Service Level Indicators)
• Set up automated remediation
• Create SLO dashboards for stakeholders
• Establish SLO review processes

About Vsceptre

Vsceptre specializes in observability solutions and DevOps best practices. Our team of experts helps organizations implement robust monitoring, feature management, and application reliability solutions to minimize downtime and enhance user experience.

For further information, contact Vsceptre at charliemok@vsceptre.com