Don't be Afraid of CSI: How Critical TechWorks Orchestrates Stateful Apps with Nomad & a CSI Process
summary
Critical TechWorks builds an orchestration pipeline that uses Nomad 0.11 with CSI plus Terraform and Consul to run stateful apps (Jenkins, SonarQube, Nexus) backed by Azure Disk CSI driver. The demo provisions four VMs, registers two managed disks, deploys apps each with two Postgres DBs, then drains a client to show volumes detach, reattach and data persist. It works end-to-end but the Azure driver sometimes needs manual recovery and team contributes fixes and doc patches.
AI-written summary · about
chapters (27)
- 0:00Intro
- 0:33What we did
- 1:14Mission
- 1:38What is Nomad
- 2:06Stateful Containers
- 2:43CSI
- 4:33Host Volumes
- 5:37Container Storage Interface
- 6:24Nomad Documentation
- 6:37Results
- 7:41Successes
- 9:44Welcome
- 10:08Azure Resource Group
- 10:37Nomad States
- 11:17Overview of Process
- 12:36Starting Applications
- 14:03Running the App
- 14:24Saving the Data
- 14:41Testing
- 14:55Maintenance Window
- 15:40Drainage Train
- 15:56Redeploy Job
- 16:19Wait for Services
- 16:51Check for Updates
- 17:37Hashtags
- 17:53Conclusion
- 18:07Questions
related talks
37mDistributed Flexibility: Nomad and Vault in a Post-Kubernetes World
21mKeynote: NSG Data Analytic Architecture Service
40mNomad: Scheduling Applications at Scale
01h 08mDystopia as a Service
46mProgressive Deployments with Consul Service Mesh
38mState is Hard: An SDK for Building Stateful Applications
26mRethinking Cloud-Native Patterns on Kubernetes
33mDistributed Scheduler Hell