自分史上一番早い2024振り返り〜コロナ後、仕事は通常ペースに戻ったか〜 by IoT fullstack engineer
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
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)
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.
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.
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
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)
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
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
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
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
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.
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