Upload
dirk-wallerstorfer
View
296
Download
3
Embed Size (px)
Citation preview
How monitoring OpenStack can positively affect your sleeping habits and hairstyleDirk Wallerstorfer OpenStack Day Seattle, Sep 30th 2016
Technology Lead OpenStack
- tech enthusiast- husband- father- Austrian- never seen “Sound Of Music” - yes, I own a lederhosen - no, I don’t know how to yodel
@wall_dirk
• Learning how to OpenStack• Three node cluster•Different configurations• Troubleshooting is hard•Production?!?
•We monitor everything•Desire for transparency• Forecasting
https://openclipart.org/image/2400px/svg_to_png/219371/You-Are-Being-Monitored.png
•Main drivers• Save money• Increase Operational efficiency• Innovate, deploy apps faster
• “What is DevOps?”•Cultural change
•Day 2•Challenges•Cloud platform insights•OS = micro services• Scale•Dynamics
•Operations monitoring
https://openclipart.org/image/2400px/svg_to_png/219371/You-Are-Being-Monitored.png
• Log Management• ELK, Splunk, sumologic,
fluentd, ...• System Monitoring• Nagios, Icinga, Sensu, Zabbix,
Prometheus, Zenoss, AppFormix, ...
• Combined and more•Monasca, DataDog, Dynatrace, ...
https://collegetraxx.com/wp-content/uploads/2016/06/possibilities-sign.jpg
https://wiki.openstack.org/wiki/Operations/Toolshttps://wiki.openstack.org/wiki/Operations/Monitoring
Log Management• ELK stack et al.•Many, many log files•Alerting
System Monitoring•Nagios et al.•Resource utilization•Check system status regularly and update UI•Agent || polling data
•Alerting•OK, Warning, Critical
Seamless Monitoring for Mesos ClustersDrew Gassaway, MesosCon 2016
http://schd.ws/hosted_files/mesosconna2016/98/SeamlessMonitoringForMesosClusters.pdf
https://xkcd.com/1319/
https://xkcd.com/1319/
'Automating' comes from the roots 'auto-' meaning 'self-', and 'mating', meaning 'screwing'.
•Alerting• Thresholds• Flood of alerts
http://dipettamortgage.com/wp/wp-content/uploads/2013/07/real-estate-statistics.jpg
http://ruthe.de/archiv/632/datum/asc/
•Right tool?•App insights + OpenStack•Who is slow?• Your app will fail!•More challenges ...
•Dev – Ops •DevOps
http://cdn.coresites.factorymedia.com/dirt_new/wp-content/uploads/2010/11/foureyes.jpg
•Dev – Ops •DevOps•DevOps – Ops•DevOps – DevOps•DevOps?
•Multitenancy?•How do you do it?
http://cdn.coresites.factorymedia.com/dirt_new/wp-content/uploads/2010/11/foureyes.jpg
http://starecat.com/sure-glad-the-hole-isnt-at-our-end-sinking-boat/
Ops
DevOps
•No selective perception• See the whole thing• Single pane of glass•War room suitable
http://1.bp.blogspot.com/-jvg11-6qeEY/ValpTTZ6WMI/AAAAAAAAAEg/FmstzA2auRQ/s1600/TrueTruthGraphic.jpg
•Holistic overview•De facto standard: Cloud•Applications•User experience
https://pixabay.com/en/mathematics-formula-physics-school-989121/
Deployments are no longer static
7:00 a.m.Low Load and Service runningon minimum redundancy
12:00 a.m.Scaled up service during peak loadwith failover of problematic node
7:00 p.m.Scaled down again to lower loadand move to different geo location
You don’t fly by hand here
820 Billion dependencies
Network ProblemMushroom cloud effect