medium: From Critical User Journey to SLO/SLIs - json

cuj.json

{
    "name": "Being able to search new books",
    "actions": [
        {
            "name": "Web Service checks with ID Service to ensure User is allowed to search",
            "dependencies": ["ID Service"],
            "slos": [
                {
                    "name": "ID Service",
                    "type": "Availability",
                    "metric": "99.9% of requests complete with less than 500 code response",
                    "metric_source": "Log messages"
                },
                {
                    "name": "ID Service",
                    "type": "Latency",
                    "metric": "99.99% of requests complete in 300ms or less",
                    "metric_source": "Log messages"
                }
            ]
        },
        {
            "name": "Web Service sends search term to Search Service",
            "dependencies": [],
            "slos": [
                {
                    "name": "Search Service",
                    "type": "Availability",
                    "metric": "99.5% of requests complete with less than 500 code response",
                    "metric_source": "Service Mesh"
                },
                {
                    "name": "Search Service",
                    "type": "Latency",
                    "metric": "99% of searches complete in 1s or less",
                    "metric_source": "Service Mesh"
                }
            ]
        },
        {
            "name": "Search Service checks cache for results",
            "dependencies": ["cache"],
            "slos": [
                {
                    "name": "Search Service completes from Cache",
                    "type": "Latency",
                    "metric": "99% of Search with cache hits complete in 400ms or less",
                    "metric_source": "Custom Prometheus metric"
                },
                {
                    "name": "Cache retrieval",
                    "type": "Latency",
                    "metric": "99% of cache requests complete in 300ms or less",
                    "metric_source": "Custom Prometheus metric"
                }
            ]
        },
        {
            "name": "If not in cache Search Service queries Database (and adds results to the cache)",
            "dependencies": [
                "database"
            ],
            "slos": [
                {
                    "name": "Search Service completes from DB",
                    "type": "Latency",
                    "metric": "99% of Search from DB complete in 1s or less",
                    "metric_source": "Custom Prometheus metric"
                },
                {
                    "name": "Database retrieval",
                    "type": "Latency",
                    "metric": "99% of DB requests complete in 600ms or less",
                    "metric_source": "Custom Prometheus metric"
                }
            ]
        },
        {
            "name": "Web Service displays search results",
            "dependencies": [],
            "slos": [
                {
                    "name": "Whole CUJ completes",
                    "type": "Availability",
                    "metric": "99% of searches complete with less than 500 code response",
                    "metric_source": "Load balancer logs"
                },
                {
                    "name": "Whole CUJ completes",
                    "type": "Latency",
                    "metric": "99.5% of searches complete in 2s or less. 99% of searches complete in 1s or less",
                    "metric_source": "Load balancer logs"
                }
            ]
        }
    ]
}