Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
Building Dashboards as a Hobby
Search
Sohei Iwahori
April 24, 2024
Technology
0
950
Building Dashboards as a Hobby
趣味のダッシュボード開発とDashboards as codeまわりのお話
Sohei Iwahori
April 24, 2024
Tweet
Share
More Decks by Sohei Iwahori
See All by Sohei Iwahori
たまに起きる外部サービスの障害に備えたり備えなかったりする話
egmc
0
570
SREのためのeBPF活用ステップアップガイド
egmc
3
4.4k
eBPFと周辺技術を利用してPHPアプリケーションコードを変更しない可視化をやってみる
egmc
1
75
PHPアプリケーションにおけるeBPFの使い所
egmc
1
1.4k
Runbookに何を書き、どのようにアラートを振り分けるか?
egmc
2
4.5k
プロダクション環境の信頼性を損ねず観測する技術
egmc
9
9.7k
practices-for-making-alerts-actionable.pdf
egmc
3
8.3k
Other Decks in Technology
See All in Technology
【Ubie】AIを活用した広告アセット「爆速」生成事例 | AI_Ops_Community_Vol.2
yoshiki_0316
1
110
Oracle Base Database Service 技術詳細
oracle4engineer
PRO
15
93k
Red Hat OpenStack Services on OpenShift
tamemiya
0
120
茨城の思い出を振り返る ~CDKのセキュリティを添えて~ / 20260201 Mitsutoshi Matsuo
shift_evolve
PRO
1
320
AIエージェントを開発しよう!-AgentCore活用の勘所-
yukiogawa
0
170
Contract One Engineering Unit 紹介資料
sansan33
PRO
0
13k
データの整合性を保ちたいだけなんだ
shoheimitani
8
3.1k
SREじゃなかった僕らがenablingを通じて「SRE実践者」になるまでのリアル / SRE Kaigi 2026
aeonpeople
6
2.5k
10Xにおける品質保証活動の全体像と改善 #no_more_wait_for_test
nihonbuson
PRO
2
310
Claude_CodeでSEOを最適化する_AI_Ops_Community_Vol.2__マーケティングx_AIはここまで進化した.pdf
riku_423
2
590
30万人の同時アクセスに耐えたい!新サービスの盤石なリリースを支える負荷試験 / SRE Kaigi 2026
genda
4
1.3k
OCI Database Management サービス詳細
oracle4engineer
PRO
1
7.4k
Featured
See All Featured
Practical Orchestrator
shlominoach
191
11k
The Invisible Side of Design
smashingmag
302
51k
Agile Leadership in an Agile Organization
kimpetersen
PRO
0
83
Improving Core Web Vitals using Speculation Rules API
sergeychernyshev
21
1.4k
Site-Speed That Sticks
csswizardry
13
1.1k
GitHub's CSS Performance
jonrohan
1032
470k
SEO Brein meetup: CTRL+C is not how to scale international SEO
lindahogenes
0
2.3k
DBのスキルで生き残る技術 - AI時代におけるテーブル設計の勘所
soudai
PRO
62
50k
Google's AI Overviews - The New Search
badams
0
910
コードの90%をAIが書く世界で何が待っているのか / What awaits us in a world where 90% of the code is written by AI
rkaga
60
42k
Money Talks: Using Revenue to Get Sh*t Done
nikkihalliwell
0
150
Measuring Dark Social's Impact On Conversion and Attribution
stephenakadiri
1
130
Transcript
Building Dashboards as a Hobby 2024/04/24 Sohei Iwahori (GREE, Inc.)
who? » Sohei Iwahori (@egmc) » GREE, Inc. » Πϯϑϥ
/ Monitoring Unit Leader / Senior Lead Engineer » ओʹήʔϜͷΠϯϑϥͱࢹγεςϜ » σʔλιʔεओʹPrometheus
࠷ۙͷΈ » systemd_exporterʹsystemd_resolved ͷϝτϦΫεΛՃ͢ΔPR͕Ϛʔδ ͞Εͨ(1݄) » ͕৽όʔδϣϯ͕ͳ͔ͳ͔ϦϦʔε ͞Εͳ͍ » ޙͷϑϦ
࠷ۙͷΈ
agenda » μογϡϘʔυͷຽओԽ » झຯͷμογϡϘʔυ࡞Ͱ͏πʔϧ » ࡞ͬͨμογϡϘʔυͷհ » ·ͱΊ
μογϡϘʔυͷຽओԽ
GrafanaͷμογϡϘʔυͷخ͠͞ » ࢥ͍͍ͭͨμογϡϘʔυΛUI্͔Β࡞Ͱ͖Δ » τϥΠΞϯυΤϥʔͷ͞ » RRDtoolͰٯϙʔϥϯυه๏ͰάϥϑΛॻ͔࣌͘Βѹతͳརศੑ
RRDTool -> Grafana + PromQL
झຯͷμογϡϘʔυ࡞Ͱ ͏πʔϧ
μογϡϘʔυ։ൃͷΤίγεςϜ » Grafana as code ৭ʑ͋Δɾɾ1 » Grafana Terraform provider
» Grafana Ansible collection » Grizzly » Grafana Crossplane provider » Kubernetes Grafana Operator » ͦͷ΄͔ScenesͳͲ 1 A complete guide to managing Grafana as code: tools, tips, and tricks Ishan Jain.
ͲΕʹ͢Δ͔ » ਖ਼ͦΜͳʹ͜Ε͕͍͍ʂΈ͍ͨͳͷͳ͍ؾ͕͢ΔͷͰϒϩάͷ௨Γطଘͷ ίʔυࢿ࢈ʹ߹ΘͤͯͰ » τϥΠΞϯυΤϥʔखಈͰɺظతʹ࡞ΓࠐΈ͍ͨ߹ʹίʔυԽ » ͱ͍͑JSONܗࣜͳΒ݁ہϕʔεΛखͰ࡞ΔɺͰ » ݸਓͰ͓खܰʹΔʹखಈ࡞+Ұ෦GrizzlyΛར༻
Grizzly » https://github.com/grafana/grizzly » Grafana։ൃͷcliπʔϧʢgrrίϚϯυʣ » ݱࡏDashbaordཧ͚ͩͰͳ͘σʔλιʔεɺGrafana Cloud prometheus ͷrules/alertsɺSynthetic
MonitoringͷཧʹରԠ » v0.4.1࣌ » جຊͷه๏YAMLɺJSONɺJsonnetʢ+Grafonnetʣ͔ͭ͑Δ
Grizzyͱͷग़ձ͍ » 2020ࠒ0.0.1ͷalpha࣌ͰJsonnet(+Grafonnet)Ͱͷར༻ΨΠυ͕͋Γɺ͍࢝Ίͨ » લड़ͷϦΞϧλΠϜөػೳͳͲ͍উख͕ྑ͔ͬͨͷͰ͜Εͳ͔ͳ͔͍͍ͷͰͱ ࢥ͍ͬͯͨ » جຊGrafonnetͰදݱ͠ɺgrafanaDashboards:: ԼʹμογϡϘʔυΦϒδΣΫτ Λஔͯ͠apply͢ΔͱμογϡϘʔυʹө͞ΕΔ
» grr watchͰϑΝΠϧͷมߋݕͯ͠ଈ࠲ʹremoteͷGrafanaΠϯελϯεʹөͳͲɺ ։ൃମݧ͕Α͔ͬͨ » ͦͯ͠2021-02-27 v0.1.0͕ϦϦʔε͞Εͨ
Grizzy 0.1.0 » ٸʹه๏͕k8sελΠϧͷYAMLʹͳͬͯhidden element͕ඇਪͱͳͬͨɾɾ
Jsonnet + GrafonnetͰҾ͖ଓ͖͑·͢ (0.4.1ݱࡏ) » dashboardఆٛΛspecʹೖΕΔܗͰJsonnetͷ··ͰOK local g = import
'github.com/grafana/grafonnet/gen/grafonnet-latest/main.libsonnet'; local dashboard = g.dashboard.new("my dahsboard"); ... { apiVersion: 'grizzly.grafana.com/v1alpha1', kind: 'Dashboard', metadata: { name: "dashboard name", folder: "dashboard folder" }, spec: dashboard }
grr serve(from 0.4.0) » ϩʔΧϧͰGrafanaཱͯͯϦϞʔτͷGrafanaΛ มߋ͢Δ͜ͱͳ͘μογϡϘʔυฤू͕Ͱ͖Δ » ϦΫΤετϓϩΩγ͞ΕΔͷͰϦϞʔτͷ σʔλιʔε͕͑Δ »
UI্Ͱอଘ͢ΔͱϩʔΧϧͷμογϡϘʔυ yaml͕ߋ৽͞ΕΔʢ!?ʣ » ͱΓ͋͑ͣܗΛ࡞ͬͯɺμογϡϘʔυΛखಈ ฤूͭͭ͠όʔδϣϯཧ͢ΔɺͳͲ͕Ͱ͖· ͢ɺศར
GrizzlyͷΠϚΠνͳ » Τϥʔϝοηʔδ͕ෆ » GrafanaͷΫϨσϯγϟϧ͕ෆ͍ͯ͠Δঢ়ଶͰ invalid memory address ͱ͔ੜͬΆ͍Τϥʔ͕ग़ͨΓ »
എܠͱͯ͠࠷ۙ։ൃׂ͕ͱεϐʔσΟʔʹਐΜͰͦ͏ɺ໘༷·ͩ҆ఆ͠ ͯͳ͍ » υΩϡϝϯτ͕؆ૉͰॳखͷ͍ํ͕Α͘Θ͔Βͳ͍ » ຊޠใ૿͍͖͍ͯͨ͠Ͱ͢Ͷ
࡞ͬͨμογϡϘʔυͷ հ
࡞ͬͨμογϡϘʔυͷհ » systemd_resolved » sysload » process dashbaord with treemap
࡞ͬͨμογϡϘʔυͷհ » systemd_resolvedʢ͜Εͱʣ » sysload » process dashbaord with treemapʢ͜ΕͷΛগ͠ʣ
systemd_resolved https://grafana.com/grafana/dashboards/15816-systemd-resolved-exporter/
systemd_resolved » ಄Ͱͨ͠Έͷճऩ » systemtd_resolvedͷϝτϦΫεΛऔΔͨΊʹಠཱͨ͠exporterΛ2022ʹॻ͍ͨ » https://github.com/egmc/systemd_resolved_exporter » ࣌Λܦͯcommunityϝϯςͱͳͬͨsystemd_exporterʹಉ༷ͷ࣮Λߦ͏PRΛ ೖΕͯϚʔδ͞Εͨ
» σϞ༻ʹμογϡϘʔυ࡞͚ͬͨͲ·ͩʢsystemd_exporterͷํͰʣ͑ͳ ͍ɺϦϦʔε͞ΕͨΒྑ͔ͬͨΒͬͯΈ͍ͯͩ͘͞ɾɾ
process dashbaord with treemap https://grafana.com/grafana/dashboards/13882-process-exporter-dashboard-with-treemap/
process dashbaord with treemap » DatadogͷμογϡϘʔυʹ͋ΔϝϞϦϚοϓతͳλΠϧදࣔΛߦ͏ෳ߹άϥϑ͕Α͔ͬͨͷͰਅࣅͯ࡞ͬͨ » process-exporeter2ͷϝτϦΫεΛදࣔ͢Δ » /proc
ҎԼͷใΛͱʹ » ҙʹάϧʔϐϯάͨ͠ϓϩηε͝ͱͷϝϞϦ/cpu/io/uptimeͳͲ » ݸਓతʹಛʹ͍͍ͨͷϝϞϦͷ༻ঢ়گͱͦͷมԽ » ͳ͔ͳ͔ศརʹ͑ͨͷͰࣗࣾͷڥʹ࣋ͪࠐΜͰ͍ͬͯΔ 2 process-exporter(https://github.com/ncabatoff/process-exporter)
ϝϞϦϦʔΫͷՄࢹԽͱ͔
OOMൃੜલͷঢ়گ֬ೝͱ͔
μογϡϘʔυ ͬͱڞ༗͞Εͯ΄͍͠ » ެࣜαΠτͷํগ͠վળ͞Ε͖ͯͨ » ΧςΰϦ·ΘΓ͏ͪΐ͍΄͍͠ » ͱ͍͑ݟ͚ͭΔͷେมͳͷͰڞ༗ɺ ൃ৴͕૿͑Δͱخ͍͠ »
͍͖ͬͯ·͠ΐ͏
·ͱΊ
·ͱΊ » μογϡϘʔυ։ൃָ͍͠ » ͕ɺ͋·Γ։ൃɺڞ༗ࣄྫ͕ྲྀΕͯ͜ͳ͍ʢؾ͕͢ΔʣɺͷͰڞ༗͍͖ͯ͠ ·͠ΐ͏ » πʔϧ৭ʑ͋Δ͕ɺݸਓతʹ·ͣखಈ࡞Ͱ » Grizzlyෳͷ࡞๏Ͱ͑ΔͷͰ͕͠ΒΈ͕ͳ͚Ε͓͢͢ΊͰ͢
Thank you for listening