Cloud Operations II

Systems via Service Providers

- Azure DevOps
+ Build
+ Release
++ Deployment Groups
++ Agent Pools
++ Integration with other systems
+ VSTS
+ Issue log, Project planning, reporting tools

- AWS CodeBuild
+ Build and test source code
+ Create and upload artifacts
+ Forward/Notify other services

- AWS CodeDeploy
+ EC2 Instances
+ AWS CLI friendly
+ Avoids downtime

- AWS CodeCommit
+ Version Control
+ Trigger builds

* Note on Microservices and Functions

Azure Functions and AWS Lambda
- Azure Language
- AWS Language

- Real Time Processing
+ ETL (variety of services)
+ Application activity, log parsing, telemetry
+ Transcoding
+ Indexing
+ Integration into other services
+ API

Monitoring
- DataDog
+ Monitoring as a Service
++ Pay per agent
++ Pay per million log events
++ Pay per 10,000 test
+ Based on ELK stack
+ Support (Remote and On-Site)
+ Ease of management (multiple environments)
+ Tons of integration (AWS, PagerDuty, Azure, Active Directory)
+ Lack of documentation
- ELK
+ Three services: Elasticsearch, Logstash, Kibana
+ Separate instances or all in one
+ Elasticsearch cluster
++ Shared data (shards)
++ Minimum nodes online
++ Forwarding node (very bizarre, in my opinion)
+ Logstash Grok Parsing
++ Ruby!
++ Separate files for ease of configuration
++ Great logs
+ Kibana
++ Integration with AD/LDAP
++ Rules/Policy Based
++ Dashboards, dashboards, dashboards!
+ Dramatic differences per version
+ Great support (when they answer)
- Nagios/Sensu/Zabbix
+ Master server
+ Configure agents
+ Create rules for monitoring
+ Setup alerting