Diese Präsentation wurde erfolgreich gemeldet.
Wir verwenden Ihre LinkedIn Profilangaben und Informationen zu Ihren Aktivitäten, um Anzeigen zu personalisieren und Ihnen relevantere Inhalte anzuzeigen. Sie können Ihre Anzeigeneinstellungen jederzeit ändern.
Nicolas Steinmetz - CérénIT
Sustainable
Observability
From Bare Metal
To k8s world
© 2020 InfluxData. All rights reserved. 2
$ whoami?
∙Freelance at CérénIT: Architecture, Automation & Time Series
∙TIG & TI...
And why the TICK/TIG stack ?
Some principles to start with...
How did I get
there ? ∙Custom metrics for home sensors &
extended to platform monitoring
∙Best of breed platform
∙Nice UI ...
∙Monitoring outside the platform
∙Telegraf to collect and send metrics
to InfluxDB
∙Push mechanism
∙Grafana for Alerting an...
It’s a journey…
Once upon a time...
From bare metal to containers...
∙(Ab)use of /etc/telegraf/telegraf.d
∙Automated by Infrastructure as
Code
∙Host and application metrics 🤩
∙Docker inputs plugin provides only
general metrics (mem, cpu, net,
healthcheck)
∙Lost visibility on what happens
inside t...
∙Add telegraf in docker ! 😌
∙Get metrics back again for services
∙But not the perfect solution…
∙A new pattern is rising… 😏
To infinity and beyond...
From containers to kubernetes...
Monitoring in
k8s world ?
∙Nodes: master(s) & workers
∙Kubernetes Core Services (etcd,
coredns, etc)
∙Application and rela...
∙De facto standard
∙Ecosystem relies on prometheus
(exporters)
∙Prometheus Operator
∙Node metrics
∙Core service metrics
∙A...
∙Already have TIG !
∙Not another / custom kubernetes
tool
∙Long term storage ?
∙Analytics ?
∙Not embedded in the cluster n...
Can we get the same with
TIG/TICK Stack ?
∙Do we want to have exactly the
same data or something similar ?
∙Just not try to duplicate Prom
Operator dashboards but q...
∙Contributed an updated version of
telegraf-ds helm chart 💪
∙Mix of traditional plugins +
kubernetes input plugin
∙Opinion...
∙Inspired from Prometheus Operator
Dashboards
∙Reproduced the one that interested
me and extended them
∙Most of the metric...
∙Telegraf helm chart
∙kube_inventory input plugin
∙Prometheus input plugin
∙All inputs plugin for your services
∙Telegraf....
∙Telegraf-operator (alpha) chart
∙Inject a telegraf container as sidecar
pod
∙Telegraf classes to define
configuration to ap...
∙Young and promising initiative
∙Ready to use & Dashboard as code
∙Kubernetes dashboards by
bonitoo.io
∙Based on kubernete...
∙Take only node-exporter,
kube-state-metrics from
Prometheus Operator
∙Use telegraf to collect prometheus
metrics via prom...
Next steps and take away
∙Explore deeper InfluxDB 2.0,
especially to dissociate Alerting from
Dashboards
∙Explore deeper telegraf-operator for
in po...
∙Possible to monitor kubernetes
platform with telegraf
∙Don’t need to deploy prometheus in
your cluster
∙Leverage promethe...
Thanks for watching !
© 2020 InfluxData. All rights reserved. 29
Nicolas Steinmetz - CérénIT
Personal : @nsteinmetz
Company : @_CerenIT ; https:/...
© 2020 InfluxData. All rights reserved. 30
Nicolas Steinmetz [CérénIT] | Sustain Your Observability from Bare Metal TICK Stack and Apps to a Kubernetes World | Influ...
Nicolas Steinmetz [CérénIT] | Sustain Your Observability from Bare Metal TICK Stack and Apps to a Kubernetes World | Influ...
Nächste SlideShare
Wird geladen in …5
×

Nicolas Steinmetz [CérénIT] | Sustain Your Observability from Bare Metal TICK Stack and Apps to a Kubernetes World | InfluxDays Virtual Experience London 2020

59 Aufrufe

Veröffentlicht am

When moving your apps to Kubernetes, you need to keep your existing observability at the same level or better. Kubernetes will give you some challenge, as you can’t strictly deploy the TICK Stack as you did before, but also allow some opportunities. The talk is about my journey on this topic and will cover Telegraf as DaemonSet to fetch nodes resources, as a deployment to fetch metrics from different endpoints and hopefully with Telegraf as an operator to illustrate sidecar deployment. All these metrics will be pushed to InfluxDB (v1/v2) and may be visualized in Chronograf or Grafana.

Veröffentlicht in: Technologie
  • Als Erste(r) kommentieren

Nicolas Steinmetz [CérénIT] | Sustain Your Observability from Bare Metal TICK Stack and Apps to a Kubernetes World | InfluxDays Virtual Experience London 2020

  1. 1. Nicolas Steinmetz - CérénIT Sustainable Observability From Bare Metal To k8s world
  2. 2. © 2020 InfluxData. All rights reserved. 2 $ whoami? ∙Freelance at CérénIT: Architecture, Automation & Time Series ∙TIG & TICK user for 5 years (InfluxDB 0.9.x) ∙InfluxAce for France since 2019 ∙Host of the Paris Time Series Meetup (ptsm.io) ∙Co-Host of the BigDataHebdo(.com) podcast
  3. 3. And why the TICK/TIG stack ? Some principles to start with...
  4. 4. How did I get there ? ∙Custom metrics for home sensors & extended to platform monitoring ∙Best of breed platform ∙Nice UI and Dashboards ∙Ready to use ∙Python API (pre-telegraf world 😉) ∙Raspberry Pi compatible
  5. 5. ∙Monitoring outside the platform ∙Telegraf to collect and send metrics to InfluxDB ∙Push mechanism ∙Grafana for Alerting and Visualising My monitoring principles
  6. 6. It’s a journey…
  7. 7. Once upon a time... From bare metal to containers...
  8. 8. ∙(Ab)use of /etc/telegraf/telegraf.d ∙Automated by Infrastructure as Code ∙Host and application metrics 🤩
  9. 9. ∙Docker inputs plugin provides only general metrics (mem, cpu, net, healthcheck) ∙Lost visibility on what happens inside the container 😰 Hello Docker !
  10. 10. ∙Add telegraf in docker ! 😌 ∙Get metrics back again for services ∙But not the perfect solution… ∙A new pattern is rising… 😏
  11. 11. To infinity and beyond... From containers to kubernetes...
  12. 12. Monitoring in k8s world ? ∙Nodes: master(s) & workers ∙Kubernetes Core Services (etcd, coredns, etc) ∙Application and related kubernetes resources
  13. 13. ∙De facto standard ∙Ecosystem relies on prometheus (exporters) ∙Prometheus Operator ∙Node metrics ∙Core service metrics ∙Application metrics ∙Grafana dashboards ∙Alert Manager & Prom UI Prometheus
  14. 14. ∙Already have TIG ! ∙Not another / custom kubernetes tool ∙Long term storage ? ∙Analytics ? ∙Not embedded in the cluster nor want to enable pull monitoring from outside So why not just choose prometheus ? 🤨🤔
  15. 15. Can we get the same with TIG/TICK Stack ?
  16. 16. ∙Do we want to have exactly the same data or something similar ? ∙Just not try to duplicate Prom Operator dashboards but question your needs and the existing dashboards. ∙Available metrics may depend of your kubernetes provider Before diving !
  17. 17. ∙Contributed an updated version of telegraf-ds helm chart 💪 ∙Mix of traditional plugins + kubernetes input plugin ∙Opinionated default configuration Global & Node metrics ?
  18. 18. ∙Inspired from Prometheus Operator Dashboards ∙Reproduced the one that interested me and extended them ∙Most of the metrics are identical ; a few are different
  19. 19. ∙Telegraf helm chart ∙kube_inventory input plugin ∙Prometheus input plugin ∙All inputs plugin for your services ∙Telegraf.conf = ConfigMap Application metrics ?
  20. 20. ∙Telegraf-operator (alpha) chart ∙Inject a telegraf container as sidecar pod ∙Telegraf classes to define configuration to apply ∙Interesting for non service metrics ∙Take care of Telegraf proliferation Telegraf operator
  21. 21. ∙Young and promising initiative ∙Ready to use & Dashboard as code ∙Kubernetes dashboards by bonitoo.io ∙Based on kubernetes and kube_inventory input plugins ∙github.com > influxdata > community-templates > k8s InfluxDB Community templates
  22. 22. ∙Take only node-exporter, kube-state-metrics from Prometheus Operator ∙Use telegraf to collect prometheus metrics via prometheus input plugin A third way ?
  23. 23. Next steps and take away
  24. 24. ∙Explore deeper InfluxDB 2.0, especially to dissociate Alerting from Dashboards ∙Explore deeper telegraf-operator for in pod metrics to confirm my hypothesis Next steps
  25. 25. ∙Possible to monitor kubernetes platform with telegraf ∙Don’t need to deploy prometheus in your cluster ∙Leverage prometheus exporters with prometheus input plugin ∙Mix of Prometheus Operator and Telegraf to have best of both worlds ? ∙Watch progress of telegraf operator Take away
  26. 26. Thanks for watching !
  27. 27. © 2020 InfluxData. All rights reserved. 29 Nicolas Steinmetz - CérénIT Personal : @nsteinmetz Company : @_CerenIT ; https://www.cerenit.fr/ ; contact@cerenit.fr Paris Time Series Meetup (mostly in French) : https://www.ptsm.io/ Podcast (in French) : https://bigdatahebdo.com/ @nsteinmetz on InfluxData Slack
  28. 28. © 2020 InfluxData. All rights reserved. 30

×