Private Cloud, OpenStack, and the BMW Datacenter
summary
BMW runs about 12,000 OS images (roughly half Linux, half Windows), is over 60% virtualized and claims automation makes it 25% more cost efficient than competitors. An internal Python private-cloud project fails to scale, so a single-server Icehouse sandbox ("one room apartment") proves concepts and leads to a semi‑productive OpenStack cluster with about 16 TB storage and 100 cores. OpenStack offers a stable API/data model and no vendor license fees, but release-to-release instability remains the main concern.
AI-written summary · about
chapters (9)
- 0:00Introduction and welcome
- 1:04BMW innovation showcase
- 2:00BMW data center overview
- 3:25Standardization and efficiency
- 5:30The private cloud initiative
- 7:36Adopting openstack
- 9:43OpenStack advantages
- 11:31Challenges and future outlook
- 12:54Closing remarks
related talks
23mOpen Matters: HP & OpenStack
18mOpen Cloud: Collapsing the Layers
41mKeynote: Open Telekom Cloud
43mSupporting NASA Science with OpenStack
29mOpenStack public clouds: 7 reasons to become a local cloud service provider
20mAre the days of on-premise servers over?
36mOpen Source and the Cloud
24mPrivate cloud price-performance analysis