Make Room! Make Room!
summary
oVirt combines thin provisioning, GlusterFS, LVM and Red Hat VDO to reclaim space, compress and deduplicate VM disks so a hyper‑converged oVirt cluster can present 25x overcommit of disk while actually using little physical storage. Demo numbers: five CentOS variants run on ~4 GB physical while each reports ~250 GB logical. It works but costs CPU and ~1 GB RAM per TB for VDO index, and requires strict monitoring of guaranteed space to avoid sudden cluster-wide outages. Gluster + LVM + VDO is the stack used.
AI-written summary · about
chapters (10)
- 0:00oVirt architecture intro
- 2:58Virtualization for efficiency
- 5:48Thin provisioning in oVirt
- 9:17GlusterFS as storage backend
- 12:52Hyper-converged oVirt setup
- 15:23VDO deduplication and compression
- 22:10Over-provisioning risks and monitoring
- 27:08VDO performance and RAM overhead
- 29:23Benefits and costs summary
- 30:55Q&A on GlusterFS and monitoring
related talks
38mYour Own Little HA Cloud at the Edge (or at Home)
01h 41mBuild your own cloud with LXD
27mLessons Learned from a Large Scale OpenStack Deployment with Tripleo
46m5 architectural innovations in VMware Cloud on AWS that you can’t miss
20mFrom Agent-Based to Agentless...
20mMulti-Cloud Without the Myth: What Actually Works in Practice
40mUsing SEAPATH to Deploy Low Latency Virtual Machines with Redundancy
18mScaling Bare Metal Provisioning with Nova and Ironic at CERN