Speaker:Yuh-Jye (EJ) Chang 張裕杰
Topic: ITRI OpenStack Distribution
About Myself
• 1984-1988 NTU ME BS
• 1994-1999 Syracuse CS PhD
• 1998-2006 Lucent/Bell Labs
• 2006-2011 Alcatel-Lucent/Bell Labs
• 2011-Present ITRI/CCMA S Division
• 2015-Present ITRI/ICL F Division
Agenda
• About ITRI OpenStack Distribution
• BAMPI
• High Availability
• Disco (Cinder Plugin)
• SOFA (All flash storage)
• Peregrine (Neutron Plugin)
• PDCM (Monitoring)
Why ITRI OpenStack?
Because we need ….
• Scalable and comprehensive bare metal provisioning
• HA support for every OpenStack system component
• Standard operating procedures (SOPs) and tools for
change management
• Scalability for Internet-facing packet processing
• Overhead-minimizing network virtualization
• Physical data center administration tool
• PMLS (HaaS): Physical Machine (Hardware) Leasing
Service)
What’s inside IOD?
• Auto Deployment from Bare Metal
• ITRI OpenStack Components High Availability
• Dual Switch Protection
• Physical Data Center Monitor
• Cinder Plugin - DISCO
• Neutron Plugin - Peregrine
• Compute Node Failover
– Move VMs in the broken Host to another healthy Host
评论0
最新资源