SlideShare une entreprise Scribd logo
1  sur  23
Télécharger pour lire hors ligne
RAMCloud: Scalable
Datacenter Storage
 Entirely in DRAM

     John Ousterhout
    Stanford University
Introduction

● New research project at Stanford

● Create large-scale storage systems entirely in DRAM

● Interesting combination: scale, low latency

● The future of datacenter storage

● Low latency disruptive to database community




October 27, 2009        RAMCloud/HPTS             Slide 2
RAMCloud Overview
● Storage for datacenters
                                          Application Servers
● 10-10000 commodity
     servers
● ~64 GB DRAM/server
● All data always in RAM
● Durable and available
● High throughput:
     1M ops/sec/server
● Low-latency access:                      Storage Servers
     5-10µs RPC
                                              Datacenter
October 27, 2009          RAMCloud/HPTS                         Slide 3
Example Configurations


                                          Today        5-10 years
                   # servers              1000         1000
                   GB/server              64GB         1024GB
                   Total capacity         64TB         1PB
                   Total server cost      $4M          $4M
                   $/GB                   $60          $4




October 27, 2009                       RAMCloud/HPTS                Slide 4
RAMCloud Motivation
● Relational databases don’t scale

● Every large-scale Web application has problems:
        Facebook: 4000 MySQL servers + 2000 memcached servers

● New forms of storage starting to appear:
           Bigtable
           Dynamo
           PNUTS
           H-store
           memcached




October 27, 2009              RAMCloud/HPTS                      Slide 5
RAMCloud Motivation, cont’d
Disk access rate not keeping up with capacity:

                             Mid-1980’s       2009      Change
   Disk capacity             30 MB            500 GB    16667x
   Max. transfer rate        2 MB/s           100 MB/s 50x
   Latency (seek & rotate)   20 ms            10 ms     2x
   Capacity/bandwidth
                             15 s             5000 s    333x
   (large blocks)
   Capacity/bandwidth
                             600 s            58 days   8333x
   (1KB blocks)
   Jim Gray's rule           5 min            30 hrs    360x

● Disks must become more archival
● More information must move to memory
October 27, 2009              RAMCloud/HPTS                      Slide 6
Impact of Latency
   Traditional Application                                          Web Application




                                      Application Servers




                                                                                         Storage Servers
              UI                                             UI
           App.                                             Bus.
           Logic                                            Logic
                          Data
                       Structures
      Single machine
                                      Datacenter


       << 1µs latency                                          0.5-10ms latency

● Large-scale apps struggle with high latency
● RAMCloud goal: low latency and large scale
● Enable a new breed of information-intensive
     applications
October 27, 2009                    RAMCloud/HPTS                                     Slide 7
Research Issues

● Achieving 5-10 µs RPC

● Durability at low latency

● Data model

● Concurrency/consistency model

● Data distribution, scaling

● Automated management

● Multi-tenancy

● Node architecture
October 27, 2009         RAMCloud/HPTS   Slide 8
Low Latency: SQL is Dead?
● Relational query model tied to high latency:
        Describe what you need up fron
        DBMS optimizes retrieval

● With sufficiently low latency:
        Don't need optimization; make individual requests as needed
        Can't afford query processing overhead
        The relational query model will disappear

● Question: what systems offer very low latency and
     use relational model?




October 27, 2009                 RAMCloud/HPTS                         Slide 9
Low Latency: Stronger Consistency?

● Cost of consistency rises with transaction overlap:
                   O ~ R*D
                   O = # overlapping transactions
                   R = arrival rate of new transactions
                   D = duration of each transaction
● R increases with system scale
        Eventually, scale makes consistency unaffordable

● But, D decreases with lower latency
        Stronger consistency affordable at larger scale
        Is this phenomenon strong enough to matter?



October 27, 2009                    RAMCloud/HPTS           Slide 10
Low Latency: One Size Fits All Again?

● "One-size-fits-all is dead" - Mike Stonebraker

● Specialized databases proliferating:
        50x performance improvements in specialized domains
        Optimize disk layout to eliminate seeks

● With low latency:
        Layout doesn't matter
        General-purpose is fast
        One-size-fits-all rides again?




October 27, 2009                   RAMCloud/HPTS               Slide 11
Conclusions
● All online data is moving to RAM
● RAMClouds = the future of datacenter storage
● Low latency will change everything:
           New applications
           Stronger consistency at scale
           One-size-fits-all again
           SQL is dead

● 1000-10000 clients
     accessing 100TB - 1PB
     @ 5-10µs latency


October 27, 2009                    RAMCloud/HPTS   Slide 12
Questions/Comments?



For more on RAMCloud motivation & research issues:
        “The Case for RAMClouds: Scalable High-Performance Storage
         Entirely in DRAM”
        To appear in Operating Systems Review
        http://www.stanford.edu/~ouster/cgi-bin/papers/ramcloud.pdf
        Or, google “RAMCloud”


October 27, 2009                    RAMCloud/HPTS                      Slide 13
Backup Slides




October 27, 2009       RAMCloud/HPTS   Slide 14
Why Not a Caching Approach?
● Lost performance:
        1% misses → 10x performance degradation

● Won’t save much money:
        Already have to keep information in memory
        Example: Facebook caches ~75% of data size

● Changes disk management issues:
        Optimize for reads, vs. writes & recovery




October 27, 2009                  RAMCloud/HPTS       Slide 15
Why not Flash Memory?

● Many candidate technologies besides DRAM
        Flash (NAND, NOR)
        PC RAM
        …

● DRAM enables lowest latency:
        5-10x faster than flash

● Most RAMCloud techniques will apply to other
     technologies
● Ultimately, choose storage technology based on
     cost, performance, energy, not volatility

October 27, 2009                   RAMCloud/HPTS   Slide 16
Is RAMCloud Capacity Sufficient?
● Facebook: 200 TB of (non-image) data today
● Amazon:
       Revenues/year:            $16B
       Orders/year:              400M? ($40/order?)
       Bytes/order:              1000-10000?
       Order data/year:          0.4-4.0 TB?
       RAMCloud cost:            $24K-240K?

● United Airlines:
       Total flights/day:        4000? (30,000 for all airlines in U.S.)
       Passenger flights/year:   200M?
       Bytes/passenger-flight:   1000-10000?
       Order data/year:          0.2-2.0 TB?
       RAMCloud cost:            $13K-130K?

● Ready today for all online data; media soon
October 27, 2009                    RAMCloud/HPTS                          Slide 17
Data Durability/Availability
● Data must be durable when write RPC returns
● Unattractive possibilities:
        Synchronous disk write (100-1000x too slow)
        Replicate in other memories (too expensive)

● One possibility: log to RAM, then disk
                                  write
                            log                   log


                     DRAM         DRAM                  DRAM
           Storage                                             async, batch
           Servers
                     disk          disk                 disk


October 27, 2009                  RAMCloud/HPTS                           Slide 18
Durability/Availability, cont’d
● Buffered logging supports ~50K writes/sec./server
     (vs. 1M reads)
● Need fast recovery after crashes:
        Read 64 GB from disk? 10 minutes
        Shard backup data across 100’s of servers
        Reduce recovery time to 1-2 seconds

● Other issues:
        Power failures
        Cross-datacenter replication




October 27, 2009                 RAMCloud/HPTS       Slide 19
Low-Latency RPCs
Achieving 5-10µs will impact every layer of the system:
● Must reduce network latency:
        Typical today: 10-30 µs/switch,                        Switch

         5 switches each way
                                                           Switch    Switch
        Arista: 0.9 µs/switch:
         9 µs roundtrip                           Switch                  Switch

        Need cut-through routing,                Client                  Server
         congestion mgmt

● Tailor OS on server side:
        Dedicated cores
        No interrupts?
        No virtual memory?

October 27, 2009                  RAMCloud/HPTS                               Slide 20
Low-Latency RPCs, cont’d
● Client side: need efficient path through VM
        User-level access to network interface?

● Network protocol stack
        TCP too slow (especially with packet loss)
        Must avoid copies

● Preliminary experiments:
        10-15 µs roundtrip
        Direct connection: no switches




October 27, 2009                  RAMCloud/HPTS       Slide 21
Interesting Facets
● Use each system property to improve the others
● High server throughput:
        No replication for performance, only durability?
        Simplifies consistency issues

● Low-latency RPC:
        Cheap to reflect writes to backup servers
        Stronger consistency?

● 1000’s of servers:
        Sharded backups for fast recovery




October 27, 2009                   RAMCloud/HPTS            Slide 22
New Conference!
USENIX Conference on Web Application Development:

● All topics related to developing and deploying Web
     applications

● First conference: June 20-25, 2010, Boston

● Paper submission deadline: January 11, 2010

● http://www.usenix.org/events/webapps10/




October 27, 2009        RAMCloud/HPTS             Slide 23

Contenu connexe

Tendances

Design decision nfs-versus_fc_storage v_0.3
Design decision nfs-versus_fc_storage v_0.3Design decision nfs-versus_fc_storage v_0.3
Design decision nfs-versus_fc_storage v_0.3David Pasek
 
The ideal and reality of NVDIMM RAS
The ideal and reality of NVDIMM RASThe ideal and reality of NVDIMM RAS
The ideal and reality of NVDIMM RASYasunori Goto
 
CPU Scheduling for Virtual Desktop Infrastructure
CPU Scheduling for Virtual Desktop InfrastructureCPU Scheduling for Virtual Desktop Infrastructure
CPU Scheduling for Virtual Desktop InfrastructureHwanju Kim
 
Presentation db2 best practices for optimal performance
Presentation   db2 best practices for optimal performancePresentation   db2 best practices for optimal performance
Presentation db2 best practices for optimal performancexKinAnx
 
VMware Performance Troubleshooting
VMware Performance TroubleshootingVMware Performance Troubleshooting
VMware Performance Troubleshootingglbsolutions
 
Warehouse scale computer
Warehouse scale computerWarehouse scale computer
Warehouse scale computerHassan A-j
 
Sc19 ibm hms final
Sc19 ibm hms finalSc19 ibm hms final
Sc19 ibm hms finalYutaka Kawai
 
Hyper-V High Availability and Live Migration
Hyper-V High Availability and Live MigrationHyper-V High Availability and Live Migration
Hyper-V High Availability and Live MigrationPaulo Freitas
 
ALL ABOUT DB2 DSNZPARM
ALL ABOUT DB2 DSNZPARMALL ABOUT DB2 DSNZPARM
ALL ABOUT DB2 DSNZPARMIBM
 
Application Live Migration in LAN/WAN Environment
Application Live Migration in LAN/WAN EnvironmentApplication Live Migration in LAN/WAN Environment
Application Live Migration in LAN/WAN EnvironmentMahendra Kutare
 
Get More Out of MongoDB with TokuMX
Get More Out of MongoDB with TokuMXGet More Out of MongoDB with TokuMX
Get More Out of MongoDB with TokuMXTim Callaghan
 
LIQUID-A Scalable Deduplication File System For Virtual Machine Images
LIQUID-A Scalable Deduplication File System For Virtual Machine ImagesLIQUID-A Scalable Deduplication File System For Virtual Machine Images
LIQUID-A Scalable Deduplication File System For Virtual Machine Imagesfabna benz
 
Building a KVM-based Hypervisor for a Heterogeneous System Architecture Compl...
Building a KVM-based Hypervisor for a Heterogeneous System Architecture Compl...Building a KVM-based Hypervisor for a Heterogeneous System Architecture Compl...
Building a KVM-based Hypervisor for a Heterogeneous System Architecture Compl...Hann Yu-Ju Huang
 
3. CPU virtualization and scheduling
3. CPU virtualization and scheduling3. CPU virtualization and scheduling
3. CPU virtualization and schedulingHwanju Kim
 
Advanced performance troubleshooting using esxtop
Advanced performance troubleshooting using esxtopAdvanced performance troubleshooting using esxtop
Advanced performance troubleshooting using esxtopAlan Renouf
 

Tendances (20)

Design decision nfs-versus_fc_storage v_0.3
Design decision nfs-versus_fc_storage v_0.3Design decision nfs-versus_fc_storage v_0.3
Design decision nfs-versus_fc_storage v_0.3
 
The ideal and reality of NVDIMM RAS
The ideal and reality of NVDIMM RASThe ideal and reality of NVDIMM RAS
The ideal and reality of NVDIMM RAS
 
CPU Scheduling for Virtual Desktop Infrastructure
CPU Scheduling for Virtual Desktop InfrastructureCPU Scheduling for Virtual Desktop Infrastructure
CPU Scheduling for Virtual Desktop Infrastructure
 
Presentation db2 best practices for optimal performance
Presentation   db2 best practices for optimal performancePresentation   db2 best practices for optimal performance
Presentation db2 best practices for optimal performance
 
VMware Performance Troubleshooting
VMware Performance TroubleshootingVMware Performance Troubleshooting
VMware Performance Troubleshooting
 
Warehouse scale computer
Warehouse scale computerWarehouse scale computer
Warehouse scale computer
 
Sc19 ibm hms final
Sc19 ibm hms finalSc19 ibm hms final
Sc19 ibm hms final
 
Hyper-V High Availability and Live Migration
Hyper-V High Availability and Live MigrationHyper-V High Availability and Live Migration
Hyper-V High Availability and Live Migration
 
ALL ABOUT DB2 DSNZPARM
ALL ABOUT DB2 DSNZPARMALL ABOUT DB2 DSNZPARM
ALL ABOUT DB2 DSNZPARM
 
Memory virtualization
Memory virtualizationMemory virtualization
Memory virtualization
 
Application Live Migration in LAN/WAN Environment
Application Live Migration in LAN/WAN EnvironmentApplication Live Migration in LAN/WAN Environment
Application Live Migration in LAN/WAN Environment
 
XS Boston 2008 Memory Overcommit
XS Boston 2008 Memory OvercommitXS Boston 2008 Memory Overcommit
XS Boston 2008 Memory Overcommit
 
XS Boston 2008 Quantitative
XS Boston 2008 QuantitativeXS Boston 2008 Quantitative
XS Boston 2008 Quantitative
 
Get More Out of MongoDB with TokuMX
Get More Out of MongoDB with TokuMXGet More Out of MongoDB with TokuMX
Get More Out of MongoDB with TokuMX
 
VM Live Migration Speedup in Xen
VM Live Migration Speedup in XenVM Live Migration Speedup in Xen
VM Live Migration Speedup in Xen
 
LIQUID-A Scalable Deduplication File System For Virtual Machine Images
LIQUID-A Scalable Deduplication File System For Virtual Machine ImagesLIQUID-A Scalable Deduplication File System For Virtual Machine Images
LIQUID-A Scalable Deduplication File System For Virtual Machine Images
 
Building a KVM-based Hypervisor for a Heterogeneous System Architecture Compl...
Building a KVM-based Hypervisor for a Heterogeneous System Architecture Compl...Building a KVM-based Hypervisor for a Heterogeneous System Architecture Compl...
Building a KVM-based Hypervisor for a Heterogeneous System Architecture Compl...
 
3. CPU virtualization and scheduling
3. CPU virtualization and scheduling3. CPU virtualization and scheduling
3. CPU virtualization and scheduling
 
2. OS vs. VMM
2. OS vs. VMM2. OS vs. VMM
2. OS vs. VMM
 
Advanced performance troubleshooting using esxtop
Advanced performance troubleshooting using esxtopAdvanced performance troubleshooting using esxtop
Advanced performance troubleshooting using esxtop
 

Similaire à Scalable Datacenter Storage Entirely in DRAM

Analytics, Big Data and Nonvolatile Memory Architectures – Why you Should Car...
Analytics, Big Data and Nonvolatile Memory Architectures – Why you Should Car...Analytics, Big Data and Nonvolatile Memory Architectures – Why you Should Car...
Analytics, Big Data and Nonvolatile Memory Architectures – Why you Should Car...StampedeCon
 
20160811 s301 e_prabhat
20160811 s301 e_prabhat20160811 s301 e_prabhat
20160811 s301 e_prabhatKumar Prabhat
 
IMC Summit 2016 Breakout - Brian Bulkowski - NVMe, Storage Class Memory and O...
IMC Summit 2016 Breakout - Brian Bulkowski - NVMe, Storage Class Memory and O...IMC Summit 2016 Breakout - Brian Bulkowski - NVMe, Storage Class Memory and O...
IMC Summit 2016 Breakout - Brian Bulkowski - NVMe, Storage Class Memory and O...In-Memory Computing Summit
 
Flexible and Scalable Domain-Specific Architectures
Flexible and Scalable Domain-Specific ArchitecturesFlexible and Scalable Domain-Specific Architectures
Flexible and Scalable Domain-Specific ArchitecturesNetronome
 
Flash Storage Trends
Flash Storage TrendsFlash Storage Trends
Flash Storage TrendsOsys AG
 
HPC DAY 2017 | HPE Storage and Data Management for Big Data
HPC DAY 2017 | HPE Storage and Data Management for Big DataHPC DAY 2017 | HPE Storage and Data Management for Big Data
HPC DAY 2017 | HPE Storage and Data Management for Big DataHPC DAY
 
Diablo Memory Channel Flash Podcast
Diablo Memory Channel Flash PodcastDiablo Memory Channel Flash Podcast
Diablo Memory Channel Flash Podcastinside-BigData.com
 
SanDisk: Persistent Memory and Cassandra
SanDisk: Persistent Memory and CassandraSanDisk: Persistent Memory and Cassandra
SanDisk: Persistent Memory and CassandraDataStax Academy
 
Netflix Open Source Meetup Season 4 Episode 2
Netflix Open Source Meetup Season 4 Episode 2Netflix Open Source Meetup Season 4 Episode 2
Netflix Open Source Meetup Season 4 Episode 2aspyker
 
Presentation db2 best practices for optimal performance
Presentation   db2 best practices for optimal performancePresentation   db2 best practices for optimal performance
Presentation db2 best practices for optimal performancesolarisyougood
 
The benefits of IBM FlashSystems
The benefits of IBM FlashSystemsThe benefits of IBM FlashSystems
The benefits of IBM FlashSystemsLuca Comparini
 
Redis Developers Day 2014 - Redis Labs Talks
Redis Developers Day 2014 - Redis Labs TalksRedis Developers Day 2014 - Redis Labs Talks
Redis Developers Day 2014 - Redis Labs TalksRedis Labs
 
Storage and performance- Batch processing, Whiptail
Storage and performance- Batch processing, WhiptailStorage and performance- Batch processing, Whiptail
Storage and performance- Batch processing, WhiptailInternet World
 
2009-01-28 DOI NBC Red Hat on System z Performance Considerations
2009-01-28 DOI NBC Red Hat on System z Performance Considerations2009-01-28 DOI NBC Red Hat on System z Performance Considerations
2009-01-28 DOI NBC Red Hat on System z Performance ConsiderationsShawn Wells
 
IBM flash systems
IBM flash systems IBM flash systems
IBM flash systems Solv AS
 
Challenges with Gluster and Persistent Memory with Dan Lambright
Challenges with Gluster and Persistent Memory with Dan LambrightChallenges with Gluster and Persistent Memory with Dan Lambright
Challenges with Gluster and Persistent Memory with Dan LambrightGluster.org
 
Effectively deploying hadoop to the cloud
Effectively  deploying hadoop to the cloudEffectively  deploying hadoop to the cloud
Effectively deploying hadoop to the cloudAvinash Ramineni
 

Similaire à Scalable Datacenter Storage Entirely in DRAM (20)

Analytics, Big Data and Nonvolatile Memory Architectures – Why you Should Car...
Analytics, Big Data and Nonvolatile Memory Architectures – Why you Should Car...Analytics, Big Data and Nonvolatile Memory Architectures – Why you Should Car...
Analytics, Big Data and Nonvolatile Memory Architectures – Why you Should Car...
 
20160811 s301 e_prabhat
20160811 s301 e_prabhat20160811 s301 e_prabhat
20160811 s301 e_prabhat
 
IMC Summit 2016 Breakout - Brian Bulkowski - NVMe, Storage Class Memory and O...
IMC Summit 2016 Breakout - Brian Bulkowski - NVMe, Storage Class Memory and O...IMC Summit 2016 Breakout - Brian Bulkowski - NVMe, Storage Class Memory and O...
IMC Summit 2016 Breakout - Brian Bulkowski - NVMe, Storage Class Memory and O...
 
Flexible and Scalable Domain-Specific Architectures
Flexible and Scalable Domain-Specific ArchitecturesFlexible and Scalable Domain-Specific Architectures
Flexible and Scalable Domain-Specific Architectures
 
Flash Storage Trends
Flash Storage TrendsFlash Storage Trends
Flash Storage Trends
 
HPC DAY 2017 | HPE Storage and Data Management for Big Data
HPC DAY 2017 | HPE Storage and Data Management for Big DataHPC DAY 2017 | HPE Storage and Data Management for Big Data
HPC DAY 2017 | HPE Storage and Data Management for Big Data
 
Diablo Memory Channel Flash Podcast
Diablo Memory Channel Flash PodcastDiablo Memory Channel Flash Podcast
Diablo Memory Channel Flash Podcast
 
SanDisk: Persistent Memory and Cassandra
SanDisk: Persistent Memory and CassandraSanDisk: Persistent Memory and Cassandra
SanDisk: Persistent Memory and Cassandra
 
RAMinate Invited Talk at NII
RAMinate Invited Talk at NIIRAMinate Invited Talk at NII
RAMinate Invited Talk at NII
 
Ramcloud
RamcloudRamcloud
Ramcloud
 
Netflix Open Source Meetup Season 4 Episode 2
Netflix Open Source Meetup Season 4 Episode 2Netflix Open Source Meetup Season 4 Episode 2
Netflix Open Source Meetup Season 4 Episode 2
 
Presentation db2 best practices for optimal performance
Presentation   db2 best practices for optimal performancePresentation   db2 best practices for optimal performance
Presentation db2 best practices for optimal performance
 
The benefits of IBM FlashSystems
The benefits of IBM FlashSystemsThe benefits of IBM FlashSystems
The benefits of IBM FlashSystems
 
Redis Developers Day 2014 - Redis Labs Talks
Redis Developers Day 2014 - Redis Labs TalksRedis Developers Day 2014 - Redis Labs Talks
Redis Developers Day 2014 - Redis Labs Talks
 
Storage and performance- Batch processing, Whiptail
Storage and performance- Batch processing, WhiptailStorage and performance- Batch processing, Whiptail
Storage and performance- Batch processing, Whiptail
 
2009-01-28 DOI NBC Red Hat on System z Performance Considerations
2009-01-28 DOI NBC Red Hat on System z Performance Considerations2009-01-28 DOI NBC Red Hat on System z Performance Considerations
2009-01-28 DOI NBC Red Hat on System z Performance Considerations
 
IBM flash systems
IBM flash systems IBM flash systems
IBM flash systems
 
Challenges with Gluster and Persistent Memory with Dan Lambright
Challenges with Gluster and Persistent Memory with Dan LambrightChallenges with Gluster and Persistent Memory with Dan Lambright
Challenges with Gluster and Persistent Memory with Dan Lambright
 
Effectively deploying hadoop to the cloud
Effectively  deploying hadoop to the cloudEffectively  deploying hadoop to the cloud
Effectively deploying hadoop to the cloud
 
RAMinate ACM SoCC 2016 Talk
RAMinate ACM SoCC 2016 TalkRAMinate ACM SoCC 2016 Talk
RAMinate ACM SoCC 2016 Talk
 

Plus de George Ang

Wrapper induction construct wrappers automatically to extract information f...
Wrapper induction   construct wrappers automatically to extract information f...Wrapper induction   construct wrappers automatically to extract information f...
Wrapper induction construct wrappers automatically to extract information f...George Ang
 
Opinion mining and summarization
Opinion mining and summarizationOpinion mining and summarization
Opinion mining and summarizationGeorge Ang
 
Huffman coding
Huffman codingHuffman coding
Huffman codingGeorge Ang
 
Do not crawl in the dust 
different ur ls similar text
Do not crawl in the dust 
different ur ls similar textDo not crawl in the dust 
different ur ls similar text
Do not crawl in the dust 
different ur ls similar textGeorge Ang
 
大规模数据处理的那些事儿
大规模数据处理的那些事儿大规模数据处理的那些事儿
大规模数据处理的那些事儿George Ang
 
腾讯大讲堂02 休闲游戏发展的文化趋势
腾讯大讲堂02 休闲游戏发展的文化趋势腾讯大讲堂02 休闲游戏发展的文化趋势
腾讯大讲堂02 休闲游戏发展的文化趋势George Ang
 
腾讯大讲堂03 qq邮箱成长历程
腾讯大讲堂03 qq邮箱成长历程腾讯大讲堂03 qq邮箱成长历程
腾讯大讲堂03 qq邮箱成长历程George Ang
 
腾讯大讲堂04 im qq
腾讯大讲堂04 im qq腾讯大讲堂04 im qq
腾讯大讲堂04 im qqGeorge Ang
 
腾讯大讲堂05 面向对象应对之道
腾讯大讲堂05 面向对象应对之道腾讯大讲堂05 面向对象应对之道
腾讯大讲堂05 面向对象应对之道George Ang
 
腾讯大讲堂06 qq邮箱性能优化
腾讯大讲堂06 qq邮箱性能优化腾讯大讲堂06 qq邮箱性能优化
腾讯大讲堂06 qq邮箱性能优化George Ang
 
腾讯大讲堂07 qq空间
腾讯大讲堂07 qq空间腾讯大讲堂07 qq空间
腾讯大讲堂07 qq空间George Ang
 
腾讯大讲堂08 可扩展web架构探讨
腾讯大讲堂08 可扩展web架构探讨腾讯大讲堂08 可扩展web架构探讨
腾讯大讲堂08 可扩展web架构探讨George Ang
 
腾讯大讲堂09 如何建设高性能网站
腾讯大讲堂09 如何建设高性能网站腾讯大讲堂09 如何建设高性能网站
腾讯大讲堂09 如何建设高性能网站George Ang
 
腾讯大讲堂01 移动qq产品发展历程
腾讯大讲堂01 移动qq产品发展历程腾讯大讲堂01 移动qq产品发展历程
腾讯大讲堂01 移动qq产品发展历程George Ang
 
腾讯大讲堂10 customer engagement
腾讯大讲堂10 customer engagement腾讯大讲堂10 customer engagement
腾讯大讲堂10 customer engagementGeorge Ang
 
腾讯大讲堂11 拍拍ce工作经验分享
腾讯大讲堂11 拍拍ce工作经验分享腾讯大讲堂11 拍拍ce工作经验分享
腾讯大讲堂11 拍拍ce工作经验分享George Ang
 
腾讯大讲堂14 qq直播(qq live) 介绍
腾讯大讲堂14 qq直播(qq live) 介绍腾讯大讲堂14 qq直播(qq live) 介绍
腾讯大讲堂14 qq直播(qq live) 介绍George Ang
 
腾讯大讲堂15 市场研究及数据分析理念及方法概要介绍
腾讯大讲堂15 市场研究及数据分析理念及方法概要介绍腾讯大讲堂15 市场研究及数据分析理念及方法概要介绍
腾讯大讲堂15 市场研究及数据分析理念及方法概要介绍George Ang
 
腾讯大讲堂15 市场研究及数据分析理念及方法概要介绍
腾讯大讲堂15 市场研究及数据分析理念及方法概要介绍腾讯大讲堂15 市场研究及数据分析理念及方法概要介绍
腾讯大讲堂15 市场研究及数据分析理念及方法概要介绍George Ang
 
腾讯大讲堂16 产品经理工作心得分享
腾讯大讲堂16 产品经理工作心得分享腾讯大讲堂16 产品经理工作心得分享
腾讯大讲堂16 产品经理工作心得分享George Ang
 

Plus de George Ang (20)

Wrapper induction construct wrappers automatically to extract information f...
Wrapper induction   construct wrappers automatically to extract information f...Wrapper induction   construct wrappers automatically to extract information f...
Wrapper induction construct wrappers automatically to extract information f...
 
Opinion mining and summarization
Opinion mining and summarizationOpinion mining and summarization
Opinion mining and summarization
 
Huffman coding
Huffman codingHuffman coding
Huffman coding
 
Do not crawl in the dust 
different ur ls similar text
Do not crawl in the dust 
different ur ls similar textDo not crawl in the dust 
different ur ls similar text
Do not crawl in the dust 
different ur ls similar text
 
大规模数据处理的那些事儿
大规模数据处理的那些事儿大规模数据处理的那些事儿
大规模数据处理的那些事儿
 
腾讯大讲堂02 休闲游戏发展的文化趋势
腾讯大讲堂02 休闲游戏发展的文化趋势腾讯大讲堂02 休闲游戏发展的文化趋势
腾讯大讲堂02 休闲游戏发展的文化趋势
 
腾讯大讲堂03 qq邮箱成长历程
腾讯大讲堂03 qq邮箱成长历程腾讯大讲堂03 qq邮箱成长历程
腾讯大讲堂03 qq邮箱成长历程
 
腾讯大讲堂04 im qq
腾讯大讲堂04 im qq腾讯大讲堂04 im qq
腾讯大讲堂04 im qq
 
腾讯大讲堂05 面向对象应对之道
腾讯大讲堂05 面向对象应对之道腾讯大讲堂05 面向对象应对之道
腾讯大讲堂05 面向对象应对之道
 
腾讯大讲堂06 qq邮箱性能优化
腾讯大讲堂06 qq邮箱性能优化腾讯大讲堂06 qq邮箱性能优化
腾讯大讲堂06 qq邮箱性能优化
 
腾讯大讲堂07 qq空间
腾讯大讲堂07 qq空间腾讯大讲堂07 qq空间
腾讯大讲堂07 qq空间
 
腾讯大讲堂08 可扩展web架构探讨
腾讯大讲堂08 可扩展web架构探讨腾讯大讲堂08 可扩展web架构探讨
腾讯大讲堂08 可扩展web架构探讨
 
腾讯大讲堂09 如何建设高性能网站
腾讯大讲堂09 如何建设高性能网站腾讯大讲堂09 如何建设高性能网站
腾讯大讲堂09 如何建设高性能网站
 
腾讯大讲堂01 移动qq产品发展历程
腾讯大讲堂01 移动qq产品发展历程腾讯大讲堂01 移动qq产品发展历程
腾讯大讲堂01 移动qq产品发展历程
 
腾讯大讲堂10 customer engagement
腾讯大讲堂10 customer engagement腾讯大讲堂10 customer engagement
腾讯大讲堂10 customer engagement
 
腾讯大讲堂11 拍拍ce工作经验分享
腾讯大讲堂11 拍拍ce工作经验分享腾讯大讲堂11 拍拍ce工作经验分享
腾讯大讲堂11 拍拍ce工作经验分享
 
腾讯大讲堂14 qq直播(qq live) 介绍
腾讯大讲堂14 qq直播(qq live) 介绍腾讯大讲堂14 qq直播(qq live) 介绍
腾讯大讲堂14 qq直播(qq live) 介绍
 
腾讯大讲堂15 市场研究及数据分析理念及方法概要介绍
腾讯大讲堂15 市场研究及数据分析理念及方法概要介绍腾讯大讲堂15 市场研究及数据分析理念及方法概要介绍
腾讯大讲堂15 市场研究及数据分析理念及方法概要介绍
 
腾讯大讲堂15 市场研究及数据分析理念及方法概要介绍
腾讯大讲堂15 市场研究及数据分析理念及方法概要介绍腾讯大讲堂15 市场研究及数据分析理念及方法概要介绍
腾讯大讲堂15 市场研究及数据分析理念及方法概要介绍
 
腾讯大讲堂16 产品经理工作心得分享
腾讯大讲堂16 产品经理工作心得分享腾讯大讲堂16 产品经理工作心得分享
腾讯大讲堂16 产品经理工作心得分享
 

Dernier

Developing An App To Navigate The Roads of Brazil
Developing An App To Navigate The Roads of BrazilDeveloping An App To Navigate The Roads of Brazil
Developing An App To Navigate The Roads of BrazilV3cube
 
A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)Gabriella Davis
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptxHampshireHUG
 
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdfThe Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdfEnterprise Knowledge
 
From Event to Action: Accelerate Your Decision Making with Real-Time Automation
From Event to Action: Accelerate Your Decision Making with Real-Time AutomationFrom Event to Action: Accelerate Your Decision Making with Real-Time Automation
From Event to Action: Accelerate Your Decision Making with Real-Time AutomationSafe Software
 
TrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law DevelopmentsTrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law DevelopmentsTrustArc
 
Factors to Consider When Choosing Accounts Payable Services Providers.pptx
Factors to Consider When Choosing Accounts Payable Services Providers.pptxFactors to Consider When Choosing Accounts Payable Services Providers.pptx
Factors to Consider When Choosing Accounts Payable Services Providers.pptxKatpro Technologies
 
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...apidays
 
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure serviceWhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure servicePooja Nehwal
 
Kalyanpur ) Call Girls in Lucknow Finest Escorts Service 🍸 8923113531 🎰 Avail...
Kalyanpur ) Call Girls in Lucknow Finest Escorts Service 🍸 8923113531 🎰 Avail...Kalyanpur ) Call Girls in Lucknow Finest Escorts Service 🍸 8923113531 🎰 Avail...
Kalyanpur ) Call Girls in Lucknow Finest Escorts Service 🍸 8923113531 🎰 Avail...gurkirankumar98700
 
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...Miguel Araújo
 
Slack Application Development 101 Slides
Slack Application Development 101 SlidesSlack Application Development 101 Slides
Slack Application Development 101 Slidespraypatel2
 
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking MenDelhi Call girls
 
Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024The Digital Insurer
 
Automating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps ScriptAutomating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps Scriptwesley chun
 
Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024The Digital Insurer
 
The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024Rafal Los
 
2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...Martijn de Jong
 
How to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerHow to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerThousandEyes
 
Workshop - Best of Both Worlds_ Combine KG and Vector search for enhanced R...
Workshop - Best of Both Worlds_ Combine  KG and Vector search for  enhanced R...Workshop - Best of Both Worlds_ Combine  KG and Vector search for  enhanced R...
Workshop - Best of Both Worlds_ Combine KG and Vector search for enhanced R...Neo4j
 

Dernier (20)

Developing An App To Navigate The Roads of Brazil
Developing An App To Navigate The Roads of BrazilDeveloping An App To Navigate The Roads of Brazil
Developing An App To Navigate The Roads of Brazil
 
A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
 
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdfThe Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
 
From Event to Action: Accelerate Your Decision Making with Real-Time Automation
From Event to Action: Accelerate Your Decision Making with Real-Time AutomationFrom Event to Action: Accelerate Your Decision Making with Real-Time Automation
From Event to Action: Accelerate Your Decision Making with Real-Time Automation
 
TrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law DevelopmentsTrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
 
Factors to Consider When Choosing Accounts Payable Services Providers.pptx
Factors to Consider When Choosing Accounts Payable Services Providers.pptxFactors to Consider When Choosing Accounts Payable Services Providers.pptx
Factors to Consider When Choosing Accounts Payable Services Providers.pptx
 
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
 
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure serviceWhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
 
Kalyanpur ) Call Girls in Lucknow Finest Escorts Service 🍸 8923113531 🎰 Avail...
Kalyanpur ) Call Girls in Lucknow Finest Escorts Service 🍸 8923113531 🎰 Avail...Kalyanpur ) Call Girls in Lucknow Finest Escorts Service 🍸 8923113531 🎰 Avail...
Kalyanpur ) Call Girls in Lucknow Finest Escorts Service 🍸 8923113531 🎰 Avail...
 
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
 
Slack Application Development 101 Slides
Slack Application Development 101 SlidesSlack Application Development 101 Slides
Slack Application Development 101 Slides
 
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
 
Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024
 
Automating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps ScriptAutomating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps Script
 
Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024
 
The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024
 
2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...
 
How to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerHow to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected Worker
 
Workshop - Best of Both Worlds_ Combine KG and Vector search for enhanced R...
Workshop - Best of Both Worlds_ Combine  KG and Vector search for  enhanced R...Workshop - Best of Both Worlds_ Combine  KG and Vector search for  enhanced R...
Workshop - Best of Both Worlds_ Combine KG and Vector search for enhanced R...
 

Scalable Datacenter Storage Entirely in DRAM

  • 1. RAMCloud: Scalable Datacenter Storage Entirely in DRAM John Ousterhout Stanford University
  • 2. Introduction ● New research project at Stanford ● Create large-scale storage systems entirely in DRAM ● Interesting combination: scale, low latency ● The future of datacenter storage ● Low latency disruptive to database community October 27, 2009 RAMCloud/HPTS Slide 2
  • 3. RAMCloud Overview ● Storage for datacenters Application Servers ● 10-10000 commodity servers ● ~64 GB DRAM/server ● All data always in RAM ● Durable and available ● High throughput: 1M ops/sec/server ● Low-latency access: Storage Servers 5-10µs RPC Datacenter October 27, 2009 RAMCloud/HPTS Slide 3
  • 4. Example Configurations Today 5-10 years # servers 1000 1000 GB/server 64GB 1024GB Total capacity 64TB 1PB Total server cost $4M $4M $/GB $60 $4 October 27, 2009 RAMCloud/HPTS Slide 4
  • 5. RAMCloud Motivation ● Relational databases don’t scale ● Every large-scale Web application has problems:  Facebook: 4000 MySQL servers + 2000 memcached servers ● New forms of storage starting to appear:  Bigtable  Dynamo  PNUTS  H-store  memcached October 27, 2009 RAMCloud/HPTS Slide 5
  • 6. RAMCloud Motivation, cont’d Disk access rate not keeping up with capacity: Mid-1980’s 2009 Change Disk capacity 30 MB 500 GB 16667x Max. transfer rate 2 MB/s 100 MB/s 50x Latency (seek & rotate) 20 ms 10 ms 2x Capacity/bandwidth 15 s 5000 s 333x (large blocks) Capacity/bandwidth 600 s 58 days 8333x (1KB blocks) Jim Gray's rule 5 min 30 hrs 360x ● Disks must become more archival ● More information must move to memory October 27, 2009 RAMCloud/HPTS Slide 6
  • 7. Impact of Latency Traditional Application Web Application Application Servers Storage Servers UI UI App. Bus. Logic Logic Data Structures Single machine Datacenter << 1µs latency 0.5-10ms latency ● Large-scale apps struggle with high latency ● RAMCloud goal: low latency and large scale ● Enable a new breed of information-intensive applications October 27, 2009 RAMCloud/HPTS Slide 7
  • 8. Research Issues ● Achieving 5-10 µs RPC ● Durability at low latency ● Data model ● Concurrency/consistency model ● Data distribution, scaling ● Automated management ● Multi-tenancy ● Node architecture October 27, 2009 RAMCloud/HPTS Slide 8
  • 9. Low Latency: SQL is Dead? ● Relational query model tied to high latency:  Describe what you need up fron  DBMS optimizes retrieval ● With sufficiently low latency:  Don't need optimization; make individual requests as needed  Can't afford query processing overhead  The relational query model will disappear ● Question: what systems offer very low latency and use relational model? October 27, 2009 RAMCloud/HPTS Slide 9
  • 10. Low Latency: Stronger Consistency? ● Cost of consistency rises with transaction overlap: O ~ R*D O = # overlapping transactions R = arrival rate of new transactions D = duration of each transaction ● R increases with system scale  Eventually, scale makes consistency unaffordable ● But, D decreases with lower latency  Stronger consistency affordable at larger scale  Is this phenomenon strong enough to matter? October 27, 2009 RAMCloud/HPTS Slide 10
  • 11. Low Latency: One Size Fits All Again? ● "One-size-fits-all is dead" - Mike Stonebraker ● Specialized databases proliferating:  50x performance improvements in specialized domains  Optimize disk layout to eliminate seeks ● With low latency:  Layout doesn't matter  General-purpose is fast  One-size-fits-all rides again? October 27, 2009 RAMCloud/HPTS Slide 11
  • 12. Conclusions ● All online data is moving to RAM ● RAMClouds = the future of datacenter storage ● Low latency will change everything:  New applications  Stronger consistency at scale  One-size-fits-all again  SQL is dead ● 1000-10000 clients accessing 100TB - 1PB @ 5-10µs latency October 27, 2009 RAMCloud/HPTS Slide 12
  • 13. Questions/Comments? For more on RAMCloud motivation & research issues:  “The Case for RAMClouds: Scalable High-Performance Storage Entirely in DRAM”  To appear in Operating Systems Review  http://www.stanford.edu/~ouster/cgi-bin/papers/ramcloud.pdf  Or, google “RAMCloud” October 27, 2009 RAMCloud/HPTS Slide 13
  • 14. Backup Slides October 27, 2009 RAMCloud/HPTS Slide 14
  • 15. Why Not a Caching Approach? ● Lost performance:  1% misses → 10x performance degradation ● Won’t save much money:  Already have to keep information in memory  Example: Facebook caches ~75% of data size ● Changes disk management issues:  Optimize for reads, vs. writes & recovery October 27, 2009 RAMCloud/HPTS Slide 15
  • 16. Why not Flash Memory? ● Many candidate technologies besides DRAM  Flash (NAND, NOR)  PC RAM  … ● DRAM enables lowest latency:  5-10x faster than flash ● Most RAMCloud techniques will apply to other technologies ● Ultimately, choose storage technology based on cost, performance, energy, not volatility October 27, 2009 RAMCloud/HPTS Slide 16
  • 17. Is RAMCloud Capacity Sufficient? ● Facebook: 200 TB of (non-image) data today ● Amazon: Revenues/year: $16B Orders/year: 400M? ($40/order?) Bytes/order: 1000-10000? Order data/year: 0.4-4.0 TB? RAMCloud cost: $24K-240K? ● United Airlines: Total flights/day: 4000? (30,000 for all airlines in U.S.) Passenger flights/year: 200M? Bytes/passenger-flight: 1000-10000? Order data/year: 0.2-2.0 TB? RAMCloud cost: $13K-130K? ● Ready today for all online data; media soon October 27, 2009 RAMCloud/HPTS Slide 17
  • 18. Data Durability/Availability ● Data must be durable when write RPC returns ● Unattractive possibilities:  Synchronous disk write (100-1000x too slow)  Replicate in other memories (too expensive) ● One possibility: log to RAM, then disk write log log DRAM DRAM DRAM Storage async, batch Servers disk disk disk October 27, 2009 RAMCloud/HPTS Slide 18
  • 19. Durability/Availability, cont’d ● Buffered logging supports ~50K writes/sec./server (vs. 1M reads) ● Need fast recovery after crashes:  Read 64 GB from disk? 10 minutes  Shard backup data across 100’s of servers  Reduce recovery time to 1-2 seconds ● Other issues:  Power failures  Cross-datacenter replication October 27, 2009 RAMCloud/HPTS Slide 19
  • 20. Low-Latency RPCs Achieving 5-10µs will impact every layer of the system: ● Must reduce network latency:  Typical today: 10-30 µs/switch, Switch 5 switches each way Switch Switch  Arista: 0.9 µs/switch: 9 µs roundtrip Switch Switch  Need cut-through routing, Client Server congestion mgmt ● Tailor OS on server side:  Dedicated cores  No interrupts?  No virtual memory? October 27, 2009 RAMCloud/HPTS Slide 20
  • 21. Low-Latency RPCs, cont’d ● Client side: need efficient path through VM  User-level access to network interface? ● Network protocol stack  TCP too slow (especially with packet loss)  Must avoid copies ● Preliminary experiments:  10-15 µs roundtrip  Direct connection: no switches October 27, 2009 RAMCloud/HPTS Slide 21
  • 22. Interesting Facets ● Use each system property to improve the others ● High server throughput:  No replication for performance, only durability?  Simplifies consistency issues ● Low-latency RPC:  Cheap to reflect writes to backup servers  Stronger consistency? ● 1000’s of servers:  Sharded backups for fast recovery October 27, 2009 RAMCloud/HPTS Slide 22
  • 23. New Conference! USENIX Conference on Web Application Development: ● All topics related to developing and deploying Web applications ● First conference: June 20-25, 2010, Boston ● Paper submission deadline: January 11, 2010 ● http://www.usenix.org/events/webapps10/ October 27, 2009 RAMCloud/HPTS Slide 23