ARM + OpenStack = ?
Jim Huang ( 黃敬群 ) Nov 10, 2016 / National Cheng-Kung University, Taiwan
Rights to copy © Copyright 2016 Jim Huang Corrections, suggestions, contributions and translations are welcome! Latest update: Nov 10, 2016 Attribution – ShareAlike 3.0 You are free ✔ to copy, distribute, display, and perform the work ✔ to make derivative works ✔ to make commercial use of the work Under the following conditions Attribution. You must give the original author credit. Share Alike. If you alter, transform, or build upon this work, you may distribute the resulting work only under a license identical to this one. ● For any reuse or distribution, you must make clear to others the license terms of this work. ● Any of these conditions can be waived if you get permission from the copyright holder. Your fair use and other rights are in no way affected by the above. License text: http://creativecommons.org/licenses/by-sa/3.0/legalcode
Goals of This Presentation • Give you an overview about – Virtualization and Computing Model – Open Source Server Community – ARM’s Server Opportunity – Virtual Private Server – OpenStack Integration • We will not discuss – Large-scale business – Profit.
Agenda
(1) Virtualization and Computing (2) Open Source Community (3) ARM cloud construction
Virtualization and Computing Model
Future Computing Trends Changes in Computing
Closed Centralized Correct Info. Stationary
Keyboard/Mouse Voice Call, SMS
Multitouch Video Call, MMS
Open Distributed Correct+Timely Infomation Mobile
Augmented Reality Gesture Interactive 3D UI Eye-Tracking Manytouch Realtime Web
Centeralized/Concentrated Known Comm. Entities
Distributed/Scattered Unknown/Utrusted Comm. Entities
Collaboration
Keyboard/M ouse Local Store
Personal Computer
Embedded IT
Sensor Network
Every Node as Both of Client/Server
Multitouch Cloud
Single-core
Multi-core
Single-core
Multi-core
UC Berkeley Sensornet Chip (TI MSP430 8MHz core, 10KB RAM)
[2009] Tiger 1GHz Single-Core Dunnington 3GHz 6-core
Privacy
Many-core Many-core
[2012] ARM 2GHz 4-core Intel 4GHz 32-core
[2017] ARM 3GHz 8-core Intel 6GHz 128-core SensorNet Chip (128MHz core, 160KB RAM)
Realtime
Source: Xen ARM Virtualization, Xen Summit Asia 2011 by Dr. Sang-bum Suh, Samsung
Virtualiztion in Historical View • 1964 - Birth of virtualization with the IBM CP series which was a test bed for the IBM S/360 system. Provided full hardware virtualization with the ability to run 14 OS instances. • 1965 – IBM begins shipping S/360 systems, the first mass production multi-purpose mainframe. First machine to use virtual memory for ‘infinite’ storage capacity. • 1970’s – IBM S/370, more of the same • 1987 – Merge/386 becomes available allowing emulation of Intel 8086 instructions on Intel 80286 & 80386 CPUs. Could run any 8086 coded OS but was typical found running Microsoft MS-DOS.
Virtualiztion in Historical View • 1997 – Virtual PC released for Macintosh • 1998 – VMware released for Windows • 1999 – Citrix Presentation Server released for Windows • 2001 – Virtual PC released for Windows; VMware Server released (first x86 server VM) • 2003 – Xen Hypervisor released (Open Source x86); MS buys Virtual PC & releases MS Virtual PC 2004 • 2005 – MS releases Virtual Server 2005 (guest machines limited to 32bit, 4GB of RAM, & 1 CPU); Intel’s VT and AMD’s AMD-V hardware virtualization added to Server and Desktop CPUs
Virtualiztion in Historical View • 2006 – VMware Server 1.0 released for free; MS Virtual Server 2005 R2 released for free; MS Virtual PC 2007 released for free; MS buys and releases SoftGrid (now called MS App-V); Amazon begins developing the first true Cloud • 2007 – VMware Server 2.0 released; VirtualBox Open Source released; Citrix acquires Xen • 2008 – VMware buys Thinstall and releases ThinApp; VMware 6.5 released, first DX9 hardware virtualization; MS releases Hyper-V for Windows 2008 (guest machines gain 64bit support, 64 GB of RAM, & 4 CPUs); First public Cloud systems come online
Virtualiztion in Historical View • 2009 – MS releases Hyper-V R2 for Windows 2008 R2 (guest machines gain CPU pooling) • 2010 – MS releases Hyper-V R2 SP1 (guest machines gain RAM pooling and DX9 hardware support); ARM announces A15 with hardware virtualization • First ARM 32-bit servers shipped in H2 2012 • First ARM 64 bit servers ship in 2014
CPU Performance 1965 - IBM S/360 – 0.1 MIPS (133,300 IPS) 1972 - IBM S/370 – 1.0 MIPS (1,000,000 IPS) 2000 - 1 GHz Intel P3 – 3,000 MIPS (3,000,000,000 IPS) 2009 - Qualcomm Snapdragon A8 – 2,000 MIPS 2010 - Intel Core i7 – 4 x 147,600 MIPS 2010 - Qualcomm Snapdragon MP – 2 x 2,500 MIPS 2011 - Qualcomm/Samsung/nVidia A9 MP – 2 x 5,000 MIPS 2012 – ARM Cortex A15 MP – 4 x 25,000 MIPS … • Geekbench: Apple 10X (10nm), single core performance more than 3300. Intel's i7-477, data single core performance about 4000 – A10X is expected to get close to i7 single core performance • • • • • • • •
Source: http://www.bestchinanews.com/Science-Technology/608.html
Server Virtualization::Benefits • Workload consolidation
– Increase server utilization – Reduce capital, hardware, power, space, heat costs
• Legacy OS support
– Especially with large 3rd-party software products
• Instant provisioning
– Easily create new virtual machines – Easily reallocate resources (memory, processor, IO) between running virtual machines
• Migration
– Predicted hardware downtime – Workload balancing
Embedded Virtualization::Benefits • • • •
Workload consolidation Flexible resource provisioning License barrier Legacy software support
– Especially important with dozens or hundreds of embedded operating systems, commercial and even home-brew
• Reliability • Security
• (1) Hardware Consolidation
Why?
– Application Processor and Baseband Processor can share multicore ARM CPU SoC to run both Linux and RTOS efficiently. -
• (2) OS Isolation
– important call services can be effectively separated from downloaded third party applications by virtualized ARM combined with access control.
1
3
2
를
V - Core
V - Core
Core
Memory
V - Core
( Realtime Hypervisor) V - Core
Core
Multi-core
V - Core
Core
V - Core
V - Core
Core
Peripherals
AP SoC + BP SoC → Consolidated Multicore SoC
Important services
Linux 1
Linux 2
Hypervisor
Hypervisor H/W
Secure Smartphone
Linux
Virtualization SW V - Core
RTOS
Secure Kernel
General Purpose OS
Android
Hardware Rich Applications from Multiple OS
• (3) Rich User Experience
– multiple OS domains can run concurrently on a single smartphone. Source: Xen ARM Virtualization, Xen Summit Asia 2011 by Dr. Sang-bum Suh, Samsung
Nuc
Use Case: Low-cost 3G Handset • Mobile Handsets – Major applications runs on Linux – 3G Modem software stack runs on RTOS domain • Virtualization in multimedia Devices – Reduces BOM (bill of materials) – Enables the Reusability of legacy code/applications – Reduces the system development time • Instrumentation, Automation – Run RTOS for Measurement and analysis – Run a GPOS for Graphical Interface • Real cases: Motorola Evoke QA4
Hypervisor
Embedded Virtualization Use Case • • • • •
Workload consolidation Legacy software Multicore enablement Improve reliability Secure monitoring
Use Case: Workload Consolidation
Consolidate legacy systems
legacy SW legacy HW
legacy legacy legacy SW SW SW host kernel new HW
Use Case: Legacy Software Run legacy software on new core/chip/board with full virtualization legacy new legacy SW SW SW host kernel legacy HW new HW Consolidate legacy software
RT app Linux core
visualization app proprietary kernel core
RT app
visualization app proprietary kernel
Linux/KVM core core
Use Case: Multicore Enablement
Legacy uniprocessor applications legacy app legacy kernel
legacy app legacy kernel
app
app
multicore kernel host kernel
legacy app
core
core
app
core
core
data plane
data plane
host kernel core core
core
Flexible resource management data
control plane control
core
core
Use Case: Improved Reliability
Hot standby without additional hardware app app HW HW app
backup app
host kernel HW
backup app HW
Use Case: Secure Monitoring
Protect monitoring software network
network app kernel HW
app
monitor
kernel host kernel HW
Open Source Server Community
Open Source Server Community • Linaro has set up an enterprise group to maintain Linux components and tools • Existing and new members will deliver optimized core open-source software for ARM servers • Reduces costs, eliminates fragmentation, accelerates product time to market • Enables ARM Server vendors to focus on innovation and differentiated value-add
[ARM]
ARM Virtual Private Server
機會在哪？ • • • •
HA & backup 技術基本上就是 Openstack + Python debugging 架構參考日本雅虎的做法 http://www.ithome.com.tw/news/98304
• 應用 : • 小型 service 現在很多新創公司都是租用 VPS 作為公司的服務 ( 例如 Dcard 租用 Linode) • 代管，降低網路服務的營運成本 ( 不只網頁，其他網路服務一同 ) • 快速重建服務 ( 利用 snapshot) • 要求 : • HA 99% & backup, 幾乎 0 downtime, 很多 IP addr • 3. http://connect.linaro.org/resource/las16/las16-301/ • 已經大規模測試 • 自動化部署 ( 含 Linaro Patch), 整合 ceph
ARM Cloud Implementations
Reference • ARM Virtualization: CPU & MMU Issues, Prashanth Bungale, vmware • Hardware accelerated Virtualization in the ARM Cortex™ Processors, John Goodacre, ARM Ltd. (2011)
