第6回インターネットと運用技術シンポジウム WIPセッション
8 likes2,362 views
第 6 回インターネットと運用技術シンポジウム (IOTS2013) の WIP (Work In Progress) セッションの発表資料です。3 本の発表を 1 つにまとめています。3 本の内訳は以下の通り。 1. 広域分散仮想化基盤のストレージ評価の最新動向 2. 動的な広域ライブマイグレーションが可能な環境における課金モデルの検討 3. 大阪大学における仮想化基盤の設計とその増強計画
第6回インターネットと運用技術シンポジウム WIPセッション
- 1. 広域分散仮想化基盤の ストレージ評価の 最新動向 The latest results of a distributed storage system for a widely distributed virtualization infrastructure 2013/12/12 IOTS2013 WIP 柏崎 礼生 Cybermedia Center Osaka University
- 5. 1978
- 8. 80- 90
- 11. 90- 00
- 12. the Internet
- 15. 2003.8.14 Northeast blackout of 2003
- 16. in Japan
- 17. 2011.3.11 The aftermath of the 2011 Tohoku earthquake and tsunami
- 20. 石狩 Ishikari City
- 21. ２つで 十分ですよ No, four. Two, two, four.
- 22. 2011
- 24. 2012
- 25. 国立情報学研究所
- 27. 北見工大 University of the Ryukyus SINET 最長 Kitami Institute of Technology 琉球大学 Cybermedia Center Osaka University
- 29. problems
- 30. shared storage
- 31. ≒50ms
- 32. RTT > 100ms
- 33. Storage XenMotion Live Migration without shared storage > XenServer 6.1
- 37. requirement
- 38. 120000 120000 Kbytes/sec 100000 80000 60000 40000 20000 064 100000 High Random R/W Performance 80000 60000 40000 20000 0 16384 4096 1024 256 1024 409616384 65536 262144 16 1.04858e+06 4.1943e+06 File size in 2^n KBytes 1.67772e+07 4 6.71089e+07 256 64 Record size in 2^n Kbytes
- 39. POSIX準拠 interface protocl NFS, CIFS, iSCSI
- 42. Con$idential Global VM migration is also available by sharing "storage space" by VM host machines. Real time availability makes it possible. Actual data copy follows. (VM operator need virtually common Ethernet segment and fat pipe for memory copy) live migration of VM between distributed areas after Migration TOYAMA site Copy to DR-sites TOKYO site before Migration Copy to DR-sites OSAKA site Copy to DR-sites real time and active-active features seem to be just a simple "shared storage". Live migration is also possible between DR sites (it requires common subnet and fat pipe for memory copy, of course)
- 43. Con$idential Front-end servers aggregate client requests (READ / WRITE) so that, lots of back-end servers can handle user data in parallel & distributed manner. Both of performance & storage space are scalable, depends on # of servers. clients front-end (access server) back-end (core server) read blocks READ req. WRITE req. write blocks Access Gateway (via NFS, CIFS or similar) scalable performance & scalable storage size by parallel & distributing processing technology
- 45. Con$idential 1. assign a new unique ID for any updated block (to ensure consistency). 2. make replication in local site (for quick ACK) and update meta data. 3. make replication in global distributed environment (for actual data copies). back-end (multi-sites) Most important ! the key for "distributed replication" (2) create 2 copies in local for each user data, write META data, ant returns ACK (1) assign a new unique ID for any updated block, so that, ID ensures the consistency a file, consisted from many blocks (1) (1') multiplicity in multi-location, makes each user data, redundant in local, at first, 3 distributed copies, at last. (3-b) remove one of 2 local blocks, in a future. (3-b) (3-a) (3-a) make a copy in different location right after ACK. (3-a)
- 48. NFS CIFS iSCSI
- 52. ! ! ! SINET4 ! ! EXAGE / Storage ! ! RICC Copyright © 2012 Yoshiaki Kitaguchi, All right reserved.
- 54. SINET4 L2VPN, L3VPN 10Gbps 825km 829km 440km 316km 223km SINET4 417km 274km RICC Copyright © 2012 Yoshiaki Kitaguchi, All right reserved.
- 56. iozone -aceI a: full automatic mode c: Include close() in the timing calculations e: Include flush (fsync,fflush) in the timing calculations I: Use DIRECT IO if possible for all file operations.
- 57. write 120000 16384 100000 120000 Kbytes/sec 100000 60000 1024 40000 80000 256 20000 60000 0 64 40000 20000 064 16 16384 4096 1024 256 1024 409616384 65536 262144 16 1.04858e+06 4.1943e+06 File size in 2^n KBytes 1.67772e+07 4 6.71089e+07 64 64 4 256 1024 4096 1638465536 1.04858e+06 262144 4.1943e+066.71089e+07 1.67772e+07 256 Record size in 2^n Kbytes File size in 2^n KBytes Record size in 2^n Kbytes 4096 80000
- 58. write rewrite read reread 100 100 100 80 60 40 60 40 20 20 0 80 10MB 100MB 1GB 0 10GB Throughput (MB/sec) 120 Throughput (MB/sec) 120 Throughput (MB/sec) 120 100 Throughput (MB/sec) 120 80 60 40 20 10MB 100MB 1GB 0 10GB random read 60 40 20 10MB 100MB File size File size 80 1GB 0 10GB random write bkwd read 1GB 10GB record rewrite 100 100 100 80 60 40 20 0 80 60 40 20 10MB 100MB 1GB 0 10GB Throughput (MB/sec) 120 Throughput (MB/sec) 120 Throughput (MB/sec) 120 100 Throughput (MB/sec) 100MB File size 120 80 60 40 10MB 100MB 1GB 0 10GB 10MB 100MB File size stride read fwrite 1GB 10GB 100 0 Throughput (MB/sec) 100 Throughput (MB/sec) 100 20 80 60 40 20 10MB 100MB 1GB File size 10GB 0 100MB 1GB File size 10MB 100MB 1GB legend 従来方式 Exage/Storage 広域対応 Exage/Storage 80 60 40 20 10MB 0 File size 120 40 40 fread 120 60 60 File size 120 80 80 20 20 File size Throughput (MB/sec) 10MB File size 10GB 0 10MB 100MB 1GB File size 10GB 10GB
- 59. SINET4 Hiroshima University EXAGE L3VPN SINET4 Kanazawa University EXAGE L3VPN
- 60. SINET4 Kanazawa University EXAGE L3VPN SINET4 NII EXAGE L3VPN
- 62. Read (before migration) Read (after migration) Write (before migration) Write (after migration) Through put (MB/sec ) propo sed metho d shared NFS
- 67. 中川郁夫 Ikuo Nakagawa @Osaka Univ, INTEC Inc.
- 69. We have been developing a widely distributed cluster storage system and evaluating the storage along with various applications. The main advantage of our storage is its very fast random I/O performance, even though it provides a POSIX compatible file system interface on the top of distributed cluster storage.
- 70. 当初の予定
- 71. 下條真司 Shinji Shimojo @Osaka Univ, NICT
- 72. 面白く ないよね！
- 73. 本番
- 76. ブロックデータは ローカルにある
- 78. (s) 17.9 201.6 175.4 Read 400.6 Write I/O read : write 25.4 MB/s 20.9 MB/s dd 60 70MB/s(read), 50 60MB/s(write) 36
- 81. DCダウン時の DR実現検証
- 82. 来年は 米国に拠点を
- 83. Future Works
- 84. 経路最適化
- 86. 拠点間マイグレーションにおける経路最適化の実現 Layer VM L2 + L2 VPLS, IEEE802.1ad PB(Q-in-Q), IEEE802.1ah(Mac-in-Mac) L2 over L3 IP VXLAN, OTV, NVGRE ( L3 IP L3 SDN ID/Locator OpenFlow ID Locator LISP IP L4 L7 IP MAT, NEMO, MIP(Kagemusha) mSCTP SCTP DNS + Reverse NAT IP ( L2 / L3 SCTP Dynamic DNS VM Reverse NAT L2 / L3 IP 21
- 88. go to next stage
- 89. 動的な広域ライブ マイグレーションが 可能な環境における 課金モデルの検討 A Consideration of accounting model based on an availability of a dynamic wide area live migration. 2013/12/12 IOTS2013 WIP 柏崎 礼生 Cybermedia Center Osaka University
- 92. VMの所在に頓着 しなくてもいい
- 94. 課金は？
- 97. おもしろくない…
- 98. 4 cores 8GB memory 40GB storage 仮想マシン Virtualized Machines (VMs) 仮想化サーバ Virtualization Servers interface ユーザ Users クラウド事業者 Cloud Service Provider
- 99. Frequency Frequency ≒ cores Real demand Imaginary demand time 仮想マシン サービス Virtualized Machines (VMs) ユーザ IT services 4 cores 8GB memory 40GB storage Users Frequency cores time Imaginary resource cores クラウド事業者 仮想化サーバ Frequency Virtualization Servers cores time Cloud Service Provider Available supplies time ユーザがバカであればあるほど の見積もり誤りが大きいほど 儲けが大きい
- 100. Frequency Frequency Real demand time Frequency cores time Frequency Real demand Frequency cores cores Real demand time cores Frequency res Real demand cores Real demand time cores Real demand time Real demand time time
- 102. ユーザ Users サービス IT services リソース 提供 リソース 要求 仮想マシン Virtualized Machines (VMs) ユーザは複数の仮想マシン (VM)を確保し、VM上で複数 のサービスが動作する。 サービスはVMにリソースを要 求し、VMはサービスにリソー スを提供する。 Per day periodicity Frequency x cores Frequency x cores Per week periodicity Frequency x cores time (day) time (sec) Per year periodicity Frequency x cores time (sec) time (week)
- 103. 提供 要求 仮想マシン Virtualized Machines (VMs) 仮想化サーバ上で複数のVMが 動作する。 VMは仮想化サーバにリソース リソース 提供 リソース 要求 を要求し、仮想化サーバはVM にリソースを提供する。 VMが要求するリソース量はVM 仮想化サーバ Virtualization Servers が持つ複数のサービスが要求す るリソース量の累計。
- 107. Array 仮想マシン ユーザ Virtualized Machines (VMs) Users Frequency x cores サービス Require user experience Supply time (msec) Frequency IT services Per day periodicity Frequency x cores cores time (msec) Per week periodicity 仮想化サーバ Frequency x cores Virtualization Servers time (day) Per year periodicity 拠点 Frequency x cores Datacenter time (week) time
- 108. 拠点 Datacenter Virtualization Servers 仮想マシン Virtualized Machines (VMs) Migration 仮想化サーバが ダウンする時は 事前に マイグレーション 仮想化サーバ Supply Virtualization Servers 仮想マシン Virtualized Machines (VMs) 仮想化サーバ Virtualization Servers worsen UX time (msec) リソースの潤沢な Migration 仮想化サーバに 動的に マイグレーション Available supplies Frequency Maintainance Mode 仮想化サーバ Require Frequency x cores cores
- 111. Selfish routing and the price of anarchy (2006)
- 112. ジョン ナッシュ John Forbes Nash Jr. (1928∼)
- 117. 懐疑的
- 119. HPC
- 120. cloud
- 121. データセンタ 事業者との 共同研究
- 122. より強力なAPI
- 123. より知的な課金
- 124. go to next stage
- 125. 大阪大学における 仮想化基盤の設計と その増強計画 A design and a project of virtualization infrastructure in Osaka University 2013/12/12 IOTS2013 WIP 柏崎 礼生 Cybermedia Center Osaka University
- 126. 飽きた も う や め て ！
- 127. キャンパスクラウド Campus Cloud Computing Environment
- 128. 600 nm 3900 processor clock rate (MHz) 22 Intel 4004 process rule (nm) 108KHz Changes of clock rate of Intel Microprocessors
- 129. core beckton xeon westmere ivy bridge sandy bridge Dunnington Core Core 2 i7 Duo Changes of number of cores on Intel core series and Xeon processor Cybermedia Center Cyber Media Center Osaka University
- 130. サーバ集約の利点 電源設備 施設A ネットワーク 経費削減 冷房設備 経費削減 TCOの 削減 経費削減 計算機センター 経費増大 施設C 施設B Cyber Media Center Osaka University
- 131. migration
- 134. 8000 40 Number of joined organizations Number of user accounts on the campus mail system 35 7000 30 6000 25 5000 20 4000 15 3000 10 2000 5 1000 0 Number of joined organizations Number of user accounts on the campus mail system 9000 2012 April 2012 July 2012 Oct. 2013 Jan. 2013 April 2013 July 2013 Oct. 0 Cybermedia Center Osaka University
- 138. 物理コア 仮想コア : 1.7 1
- 139. averaged monthly changes of CPU usage ratio on Osaka university campus cloud system CPU usage ratio 20 15 10 5 0 2013/10/19 17:00 2013/10/24 1:00 2013/10/28 9:00 2013/11/1 17:00 2013/11/6 1:00 time 2013/11/10 9:00 2013/11/14 17:00 Cybermedia Center Osaka University
- 143. 行政の大罪 The deadly sins in public administration (1980)
- 145. (1) 高邁な目標 The first thing to do make sure that a program will not have results is to have a lofty objective.
- 146. (2) 優先順位の不在 The second strategy guarantee to produce non-performance is to try to do several things at once. It is to refuse to establish priorities and to stick to them
- 147. (3) 肥大の美学 The third deadly sin of the public administrator is to believe that fat is beautiful, despite the obvious fact that mass does not work.
- 148. (4) 根拠なき信念 Don t experiment, be dogmatic
- 149. (5) 経験に学ばず Make sure that you cannot learn from experience is the next prescription for non-perfomance in public administration.
- 150. (6) 止まらない The last of administrator s deadly sins is the most damning and the most common: the inability of abandon.
- 151. 大鑑巨砲主義
- 152. 伊四〇〇
- 157. (4) 分相応 背伸びはする 分は弁える
- 160. PLAN1 PLAN3 PLAN2 ODINS VLAN Firewalls 10GbE L2 switches Load Balancers Software Firewall Software Load Balancer L3 switches Software Router VM segment VM segment VLAN VM host segment management segment VM host segment management segment Storage Segment management segment Storage Segment management segment
- 161. VMware NSX
- 162. VMware VSAN
- 163. 目標 400仮想コアの集約
- 166. DR Disaster Recovery
- 168. ウダシティ
- 169. セバスチアン スラン Sebastian Burkhard Thrun (1967∼)
- 171. Oxford Cambridge Harvard MIT Stanford Princeton (three online Universities) Brigham Young University
- 172. Angus Maddison, The World Economy A Millennial perspective, Historical Statistics (2007)
- 173. The world s top 10 economies 1820 China India France 28.7% 16.0% 5.4% U.K. Prussia 1999 2009 24.3% 8.7% 8.6% 2050 U.S. 30.0% Japan 14.5% Germany 6.6% U.S. Japan China China U.S. India 32.7% 17.8% 17.4% 5.2% 4.9% U.K. France 4.7% 4.7% Germany 5.7% France 4.6% Brazil Mexico Japan Austria Spain 3.1% 1.9% 1.9% Italy China Spain 3.8% 3.8% 2.0% U.K. Italy Brazil 3.7% 3.6% 2.7% Russia 4.0% Indonesia 3.2% Japan 3.1% U.S. Russia 1.8% 1.7% Canada Mexico 2.1% 1.6% Spain Canada 2.5% 2.3% U.K. 2.4% Germany 2.3% 5.3% 4.3% From: BK Suh, Mega Trends: An External View , Cisco Connect 2013
- 174. 阪大のモットー motto of Osaka University
- 175. 地域に生き 世界に伸び Live locally, grow globally.
- 176. 地域に世界に 生き伸びろ Survive, locally or globally.
- 178. 転職先 探しています
- 179. おあとが よろしい ようで