ECS モニタリング手法大整理

Transcript

1. ECS ϞχλϦϯάख๏େ੔ཧ Yu Endo @ yendoooo 2025/08/21 JAWS-UG ίϯςφࢧ෦ ೖ໳ฤ #8 ॳ৺ऀେ׻ܴ LT େձ
2. Yu Endo @yendoooo1004 WHERE Inc. - 2023 Japan AWS All Certi fi cations Engineer - Former AWS Authorized Instructor - Radiological Technologist - AWS CDK - AWS Distro for OpenTelemetry (ADOT)
3. ϞχλϦϯάͷجຊཁૉ Metrics Logs Traces Alerts ਺஋ͰදݱՄೳͳଌఆ஋ Πϕϯτͷ࣌ܥྻه࿥ ෼ࢄγεςϜͰͷϦΫΤετ௥੻ ҟৗঢ়ଶͷࣗಈ௨஌ CPU ࢖༻཰: 85% Ϩεϙϯε࣌ؒ: 200ms 2024-08-18 10:30:15 ERROR: Database connection failed API → DB → Cache ͷॲཧϑϩʔ CPU ࢖༻཰ > 90% Ͱ؅ཧऀʹ௨஌
4. ECS ʹ͓͚ΔϞχλϦϯάख๏ - ECS on Fargate - Amazon CloudWatch container health checks Container Insights AWS Distro for OpenTelemetry AWS X-Ray ECS Exec FireLens Task Metadata Endpoint
5. Metrics
6. Metrics ແྉ ༗ྉ CPU ࢖༻཰ Memory ࢖༻཰ Container Insights with enhanced observability ʹΑΔৄࡉͳϞχλϦϯά 1෼ִؒ 30 ඵ ~ ਺෼ִؒ Service Ϩϕϧ Cluster - Service Ϩϕϧ Amazon CloudWatch Container Insights جຊϞχλϦϯά ίετॏࢹ Cluster / Service Ϩϕϧͷ݈શੑධՁ ΩϟύγςΟϓϥϯχϯά Task / Container ϨϕϧͷϞχλϦϯά ແྉ ϥϯλΠϜ৘ใ ϦΞϧλΠϜ Task ~ Container Ϩϕϧ Task Metadata Endpoint τϥϒϧγϡʔςΟϯά
7. Logs
8. Logs Cluster ϨϕϧͰͷઃఆ Task De fi nition Ͱͷ࠷খݶͷઃఆ Fluentd / Fluent Bit Sidecar Embedded Metric Format (EMF) ඪ४తͳΞϓϦέʔγϣϯϩά ॊೈͳϩάՃ޻ Amazon CloudWatch Container Insights FireLens γϯϓϧઃఆ جຊతͳΞϓϦέʔγϣϯϩάऩू ϩάग़ྗઌͷϧʔςΟϯά ϩάͷม׵ɾϚεΩϯά Third Party ౷߹ ΠϯϑϥɾΞϓϦ౷߹ϞχλϦϯά ύϑΥʔϚϯε෼ੳ CloudWatch Logs CloudWatch Logs ෳ਺ͷૹ৴ઌ
9. Traces
10. Traces ADOT Collector X-Ray Daemon Metrics, Logs ͱͷ౷߹؂ࢹ ৄࡉͳ Sampling rule ઃఆ AWS Distro for OpenTelemetry AWS X-Ray AWS αʔϏε౷߹ Managed Service ϚϧνΫϥ΢υ ౷߹ϞχλϦϯά ৄࡉઃఆɾΧελϚΠζ AWS Native OpenTelemetry ϕʔε
11. Alerts
12. Alerts Amazon CloudWatch Container Insights AWS αʔϏε౷߹ γϯϓϧઃఆ ৄࡉϞχλϦϯά SLI / SLO ϞχλϦϯά Container Insights Metrics ඪ४ /ΧελϜ Metrics Cluster ~ Container Ϩϕϧͷৄࡉ౓ ୯Ұ / ෳ߹ / ML ϕʔε ࠷খݶͷઃఆ Network, ύϑΥʔϚϯεϞχλϦϯά
13. Others - container health checks -
14. Others - ECS Exec -
15. ·ͱΊ Services / Tools Categories References Amazon CloudWatch Metrics, Logs, Alerts https://docs.aws.amazon.com/AmazonECS/latest/developerguide/ cloudwatch-metrics.html Task Metadata Endpoint Metrics https://docs.aws.amazon.com/AmazonECS/latest/developerguide/ task-metadata-endpoint.html Container Insights Metrics, Logs https://docs.aws.amazon.com/AmazonECS/latest/developerguide/ cloudwatch-container-insights.html FireLens Logs https://docs.aws.amazon.com/AmazonECS/latest/developerguide/ using_ fi relens.html AWS X-Ray Traces https://docs.aws.amazon.com/xray/latest/devguide/aws-xray.html AWS Distro for OpenTelemetry (ADOT) Traces https://docs.aws.amazon.com/AmazonECS/latest/developerguide/ trace-data.html container health checks Others https://docs.aws.amazon.com/AmazonECS/latest/developerguide/ healthcheck.html ECS Exec Others https://docs.aws.amazon.com/AmazonECS/latest/developerguide/ ecs-exec.html
16. Thank you !! @yendoooo1004 @yendoooo