Diving into Provisioning GlusterFS Volumes in k8s with Heketi
summary
A technical deep dive into Heketi and GlusterFS provisioning in Kubernetes that walks through the exact API calls, executors used (mock, ssh, kubectl), and on-disk steps Heketi runs (creating brick paths, mkfs.xfs, adding brick entries, and registering volumes in the JSON DB). It examines a 3-cluster, 3-brick replicated layout, documents issues like hardcoded naming, inability to auto-recreate old bricks, and heketi DB inconsistencies, then describes forking Heketi, adding custom naming, and shipping scripts to repair the heketi DB by rebuilding bricks and restoring operation.
AI-written summary · about
chapters (17)
- 0:00Introduction
- 1:25GlusterFS with Heketi in production
- 2:56What is GlusterFS
- 4:01Heketi-Kubernetes interaction
- 7:10Heketi volume creation workflow
- 12:15GlusterFS without Heketi
- 13:41Renaming GlusterFS volumes
- 15:32Heketi components: Cobra and BoltDB
- 18:45Heketi database structure
- 19:06Heketi SSH and Kubernetes executors
- 20:08GlusterFS volume issues
- 22:20Red Hat docs access
- 23:50IBM GlusterFS resources
- 24:12Volume recovery scripts
- 25:36Q&A
- 30:21Handling GlusterFS split-brain
- 33:30Customizing GlusterFS volumes
related talks
37mYou Have Stateful Apps - What if Kubernetes Would Also Run Your Storage?
19mPersistent Storage for Containers using Gluster
07mLightning Talk - Say what? You're Running the Storage Platform IN Kubernetes?
33mKubernetes As A Service with GitOps
23mConstructing Building Blocks for Production Grade Cloud Native Deployments (with Cro... Hasan Türken
35mKubernetes journey
24mDynamic Big Data in the Cloud with Kubernetes
27mKubernetes the Hard Way Using Terraform