Showing posts with label COMPUTER. Show all posts
Showing posts with label COMPUTER. Show all posts
Tuesday, May 03, 2011
Tuesday, May 18, 2010
Thursday, May 13, 2010
Cloud Computing Expo #1

Today, I went to the 1st Cloud Computing Expo (Cloud Japan) to listen to a talk given by Yuji Chino of NicoNico Video(ニコニコ動画).
Actually, it was the keynote speech of the Data Storage Expo which was joint hosted with that.
That's why, his talk was mainly focusing on storage subsystem perspective of their system.
Here are my notes:
- Nico Video has about 140TB of data in total. Individual video file has 2 or 3 copies including the original.
- At this moment, Nico Video supports only 2 levels of resolution. This is one trade off decision.
- At this moment, Nico Video does not preserve the original uploaded data. This is another trade off decision.
- Nico Video's system uses 3-tier model, edge servers(400), origin servers(100) and storage subsystem(3).
- I'm not sure what is the origin server, but it looks like a kind of cache server.
- Because of their load balancing and multi-tier cache architecture, there is no significant I/O traffic for the storage subsystem even they are using NFS, not FC fabric.
- The edge servers utilizes fast I/O throughput of SSD, for other servers capacity is primarily important.
Saturday, April 03, 2010
RFC 5841
http://www.ietf.org/rfc/rfc5841.txt
This RFC proposes a new TCP option to denote packet mood. :o
I like it. :)
This RFC proposes a new TCP option to denote packet mood. :o
I like it. :)
Friday, March 19, 2010
A LISP Textbook

LISP For Beginners Revised Edition (初めての人のためのLISP 増補改訂版)
Ikuo Takeuchi
ISBN 978-4-7981-1941-0
I bought this book today.
The author is Professor Takeuchi whom I mentioned in the beginning of this month.
Actually, I studied LISP about 15 years ago using Winston and Horn, but it's a quite standard textbook and in other word boring for me.
You can understand that the textbook is abnormal by skimming it within a minute. :o
It consists of mainly dialogs of Professor K and students, and what the most important thing is that it tells us philosophy of LISP and computer science throughout the book.
Actually, the book has been out of prints for years. But, Professor Takeuchi revised it greatly and published again yesterday. :)
It's worth reading if you are working on computer science. :)
Thursday, March 18, 2010
Bluetooth USB Adapter
Today, I went to Yurakucho BicCamera and bought the below. :)

Corega CG-BT2USB02CB
It's a Class2 Bluetooth USB adapter, and it costed me 1280 yen (roughly 14USD).
For me, the cost was the first priority this time, and the second was size.
Class 1 or 2 (maximum communication radius) was not important at all.
Look at the picture, and compare the size with a cigarette. It's so small and perfect. :)
BTW, why I bought it?
TOP SECRET. :o
Anyway, I'm going to use it at office, not at home. :)

Corega CG-BT2USB02CB
It's a Class2 Bluetooth USB adapter, and it costed me 1280 yen (roughly 14USD).
For me, the cost was the first priority this time, and the second was size.
Class 1 or 2 (maximum communication radius) was not important at all.
Look at the picture, and compare the size with a cigarette. It's so small and perfect. :)
BTW, why I bought it?
TOP SECRET. :o
Anyway, I'm going to use it at office, not at home. :)
Thursday, March 11, 2010
IPSJ 72nd National Convention, Day 3
Today, I attended at a regular session at the IPSJ national convention.
https://www.gakkai-web.net/gakkai/ipsj/72program/html/program/P5.html#6E
Here are my notes.
6E-1
Hum... Their work looked like at very early stage. I'm not
sure why they did the presentation now...
6E-2
It looks like a part of METI funded national R&D project carried on
by NEC Laboratory. The work is about a new (?) cache data control method
for resource configuration information expressed in CIM. It's a variation
of LRU resource management techniques, I think, and their basic idea
is to use residual time before expiration of various resource
configuration information. The experimental results show the validity
of the idea, but I felt it's important how to choose the TTL values of
each resource configuration information. I pointed out it after their
presentation, but
6E-3
A work for realizing multi-tenant environment without
any modifications of user Servlet applications
done by Hitachi System Development Laboratory.
The essential point of the work is that keep using a single database. :o
I would have separate databases for each tenant and just modify DBCP
(database connection pool) configuration, but it looks like
the strong constraint they assumed came from GlassFish.
Technically, their implementation stores tenant level context information
in thread local storage, intercepts some SQL statements and rewrite
table names in the single database.
Performance impact seemed to be small, but it's better to make longer
pseudo application loads than they did.
6E-4 Canceled.
6E-5
Our work on cloud federation manager which I mentioned here several times.
Actually, what we thought was that we put too much information
into our paper. :o
6E-6
An interesting work on multimedia processing system by
NTT Cyber Space Laboratory. They have been working
on a huge scale stream data processing cluster engine called DMS for
several years, and this work "Niji-Gumo" (虹雲) which means rainbow
cloud is an implementation of a new frame work supporting intuitive UI
framework using DMS as its backend engine.
The work is interesting, but it was not clear for me what theme the presenter
wants to focus on. User interface improvement, multimedia data processing or
DMS itself...?
6E-7
A work on improving a multi-tier streaming data processing engine, DMS,
from the same organization with the previous session.
DMS is basically consists of 3-tier clusters, and the second tier which
is responsible for the most CPU intensitive processing *pulls* requests
from the front tier. Furthermore, members of the second tier do not
communicate with each. This structure comes from a motivation shooting for
maximum parallelism of course. But, it naturally produces a side effect
that sometimes a lot of the second tier servers try to pull requests
from one front end server which the called 'back-pressure problem'.
This time, they chose to solve the problem making a kind of waiting queue
in the front end tier servers. :(
I thought the resolution is quite ad-hoc and they should consider the puller
side resolution too because nowadays there are good algorithms for keeping
consistency in distributed environment.
https://www.gakkai-web.net/gakkai/ipsj/72program/html/program/P5.html#6E
Here are my notes.
6E-1
Hum... Their work looked like at very early stage. I'm not
sure why they did the presentation now...
6E-2
It looks like a part of METI funded national R&D project carried on
by NEC Laboratory. The work is about a new (?) cache data control method
for resource configuration information expressed in CIM. It's a variation
of LRU resource management techniques, I think, and their basic idea
is to use residual time before expiration of various resource
configuration information. The experimental results show the validity
of the idea, but I felt it's important how to choose the TTL values of
each resource configuration information. I pointed out it after their
presentation, but
6E-3
A work for realizing multi-tenant environment without
any modifications of user Servlet applications
done by Hitachi System Development Laboratory.
The essential point of the work is that keep using a single database. :o
I would have separate databases for each tenant and just modify DBCP
(database connection pool) configuration, but it looks like
the strong constraint they assumed came from GlassFish.
Technically, their implementation stores tenant level context information
in thread local storage, intercepts some SQL statements and rewrite
table names in the single database.
Performance impact seemed to be small, but it's better to make longer
pseudo application loads than they did.
6E-4 Canceled.
6E-5
Our work on cloud federation manager which I mentioned here several times.
Actually, what we thought was that we put too much information
into our paper. :o
6E-6
An interesting work on multimedia processing system by
NTT Cyber Space Laboratory. They have been working
on a huge scale stream data processing cluster engine called DMS for
several years, and this work "Niji-Gumo" (虹雲) which means rainbow
cloud is an implementation of a new frame work supporting intuitive UI
framework using DMS as its backend engine.
The work is interesting, but it was not clear for me what theme the presenter
wants to focus on. User interface improvement, multimedia data processing or
DMS itself...?
6E-7
A work on improving a multi-tier streaming data processing engine, DMS,
from the same organization with the previous session.
DMS is basically consists of 3-tier clusters, and the second tier which
is responsible for the most CPU intensitive processing *pulls* requests
from the front tier. Furthermore, members of the second tier do not
communicate with each. This structure comes from a motivation shooting for
maximum parallelism of course. But, it naturally produces a side effect
that sometimes a lot of the second tier servers try to pull requests
from one front end server which the called 'back-pressure problem'.
This time, they chose to solve the problem making a kind of waiting queue
in the front end tier servers. :(
I thought the resolution is quite ad-hoc and they should consider the puller
side resolution too because nowadays there are good algorithms for keeping
consistency in distributed environment.
Tuesday, March 09, 2010
IPSJ 72nd National Convention

Today, I went to the 72nd National Convention of Information Processing Society of Japan and attended at a special session on cloud computing.
Here are my memorandums:
- Hadoop at Yahoo, Today and Tomorrow, Owen O'Malley, Yahoo!
- Need to look into "Oozie", a workflow engine under development at Yahoo
- Need to check "Zebra", a column oriented data store used in Pig
- Append feature of HDFS will not be included in 0.21, but under development :)
- Generating spam filter is one use case of Hadoop as machine lerning
- Cloud Computing in Government: Standards, Security, and Early Experience, Charlie Catlett, ANL
- Need to check "Magellan" project of NERSC
- Computers and Networks, Coming closer each other by cloud computing, Hideo Miyahara, NICT
- Architecture = Science + Technology (The guy stated so, but I don's subscribe to this point of view...)
- To check what is AKARI project at NICT
Monday, March 08, 2010
HHK Came Alive!

Last Friday, I wrote about my 10 years old Happy Hacking Keyboard.
I thought it was dead because of electrical trouble because I spilled green tea on it.
But, after 2 days interval, it looks like come alive. :)
I guess it's because it dried up completely, and anyway I'm feeling happy today. :)
IT Architect?
I'm running a twitter clone system at work, and I was a bit shocked seeing a tweet today. :(
Well, the guy who tweeted it is known as an really excellent engineer inside us, and his specialty covers mainly software architecture. Thus, he has been in charge of middle ware level framework such as Spring or so. Recently, he also worked on cloud computing, and thus I thought of course he knew what is GAE, Google App Engine, including Pros/Cons.
But, it turned out that he didn't know that GAE does not have support of RDBMS so far!
Of course, I believe that today he understand the essential things on GAE and do right things.
But the problem is, that even I, mainly in charge of infrastructure layer, knew such a basic issue more than 1 year ago, and he is one of lead software architects of us.
I'm wondering where my employer is going... :(
P.S.
Ohta-san, I'm not talking about Masatani-san. Other guy whom you don't know.
Well, the guy who tweeted it is known as an really excellent engineer inside us, and his specialty covers mainly software architecture. Thus, he has been in charge of middle ware level framework such as Spring or so. Recently, he also worked on cloud computing, and thus I thought of course he knew what is GAE, Google App Engine, including Pros/Cons.
But, it turned out that he didn't know that GAE does not have support of RDBMS so far!
Of course, I believe that today he understand the essential things on GAE and do right things.
But the problem is, that even I, mainly in charge of infrastructure layer, knew such a basic issue more than 1 year ago, and he is one of lead software architects of us.
I'm wondering where my employer is going... :(
P.S.
Ohta-san, I'm not talking about Masatani-san. Other guy whom you don't know.
Friday, March 05, 2010
Happy Hacking Keyboard
Wednesday, March 03, 2010
The Last Lecture of Prof. Takeuchi

Today, I attended at the last lecture of Professor Ikuo Takeuchi held at the University of Tokyo.
Actually, I was not a student nor a coworker of him, but I worked with his son for several years. So, I've been feeling strong sense of community...
The title of the lecture was "研究・開発は楽しく" which roughly means "Have Fun in Research and Development activities." Today, he walked through his surprising research activities since he was an elementary school student. According to the story, when Professor Takeuchi was 3rd grade of elementary school, he insisted that 8 divided by 8 equals! :0
Professor Takeuchi is famous for his Lisp related works such as Takeuchi Function, so
at the end of his last lecture, he quoted from Raymond M. Smullyan, "The S-expression has no purpose and for this reason fulfills all its purposes admirably."
And here is his conclusion,
"The true creator has no purpose and for this reason fulfills all its purposes admirably."
Great, isn't it... :)
Tuesday, March 02, 2010
Cloud Network Symposium

Today, I attended at a symposium titled, "Cloud Network Symposium" held in Otemachi, central Tokyo, not as just an attendee but an organizer. :o
As its subtitle "Possibilities of Secure Cloud Networking Technology" shows, the symposium was partly intended for announcing results of an R&D activity funded by the ministry of internal affairs and communications in FY 2009.
Thus, most of talks were technology oriented, but actually the most impressive speaker was Mr. Yong from Saga prefecture saying "Consider what you want/should do using cloud computing. In most of the cases, BPR is more important than technology when you try to apply new technology to your (business) system." :o
Monday, March 01, 2010
Pocket WiFi
I'm writing this entry in a moving trains car, now.
Well, I bought an EMOBILE's Pocket WiFi (D25HW) last Saturday.
D25HW is EMOBILE's new generation terminal capable of connecting to other devices such as note-pc via WiFi. As it uses conventional IEEE 801.11g I/F, it can be used with a Nintendo DS for example. Furthermore, if an iPhone user wants to access to iTunes WiFi store, D25HW can be used as iPhone's WiFi AP.
Like the picture below, it's so small and can be used without plugging into other devices. So, it's OK just power on and put it into your bag. :)

So far as connection quality, it's not so fast. But, enough for checking e-mails on your way home or office or net surfing even if you are in a train car. :)

BTW, since I got a baby 2 years ago, I've been busy both at home and at work. Especially, I had almost no time working on hobby programming at home. It's was frustrating actually.
But, I can make time on my way back home from now, and now I'm wondering which project I should work on first . :)
Well, I bought an EMOBILE's Pocket WiFi (D25HW) last Saturday.
D25HW is EMOBILE's new generation terminal capable of connecting to other devices such as note-pc via WiFi. As it uses conventional IEEE 801.11g I/F, it can be used with a Nintendo DS for example. Furthermore, if an iPhone user wants to access to iTunes WiFi store, D25HW can be used as iPhone's WiFi AP.
Like the picture below, it's so small and can be used without plugging into other devices. So, it's OK just power on and put it into your bag. :)

So far as connection quality, it's not so fast. But, enough for checking e-mails on your way home or office or net surfing even if you are in a train car. :)

BTW, since I got a baby 2 years ago, I've been busy both at home and at work. Especially, I had almost no time working on hobby programming at home. It's was frustrating actually.
But, I can make time on my way back home from now, and now I'm wondering which project I should work on first . :)
Thursday, February 11, 2010
VAIO CW
Today, we went to Yamada Denki and bought a new PC for my wife.
My wife's choice was SONY VAIO CW, and I was a bit surprised because she wanted a BluRay drive.
, 
It turned out that actually she wanted to watch "THIS IS IT" by BluRay media... :o
(It contains more bonus tracks than DVD version.) :)
BTW, I also spent my youth enjoying Micheal Jackson's tunes. So, I also enjoyed "This Is It." :)
My wife's choice was SONY VAIO CW, and I was a bit surprised because she wanted a BluRay drive.
, 
It turned out that actually she wanted to watch "THIS IS IT" by BluRay media... :o
(It contains more bonus tracks than DVD version.) :)
BTW, I also spent my youth enjoying Micheal Jackson's tunes. So, I also enjoyed "This Is It." :)
Friday, January 08, 2010
Funny Font
I found a funny font today while net surfing...
http://blog.designed-by-kenji.com/?eid=613862
I thought the designer is a genius. :o
It's a pity that it's difficult to create Japanese version of the font because kanji characters have more complicated shapes. Maybe it's possible to design a Hiragara funny font. :)
BTW, I'm wondering what if I wrote a academic paper using the font... :o
http://blog.designed-by-kenji.com/?eid=613862
I thought the designer is a genius. :o
It's a pity that it's difficult to create Japanese version of the font because kanji characters have more complicated shapes. Maybe it's possible to design a Hiragara funny font. :)
BTW, I'm wondering what if I wrote a academic paper using the font... :o
Monday, November 09, 2009
Hadoop: The Definitive Guide

Hadoop: The Definitive Guide
Tom White
ISBN 978-0-596-52197-4
Today, I made a 6 nodes small Hadoop cluster finally.
In also Japan, recently everybody and his brother of IT industory is interested in Hadoop and similar technology, and I am not an exception. :o
Around me at work, there seem to be a lot of people who indeed have interests on Hadoop but do not know what they can do using Hadoop. :o
In case of me, the situation is a bit different. I'm working on a national R&D project related to cloud computing, but actually I'm not sure if the technology is really useful or not. :o So, as a use case of it, I began to work on Hadoop/MapReduce recently. :(
I know it's like putting the cart before the horse, but what can I do?? :(
Thursday, November 05, 2009
VIMPL#20
I attended at the 20th meeting of the Virtualization Implementation Technology Workshop (Meeting?) tonight.
Today, there were 4 speakers, and three of them gave us reports of SOSP '09, and the other one talked about Japan Linux Symposium.
For me, the most interesting topic was about formal verification work of L4 microkernel(!?), seL4. Although the authors from NICTA, Australia, had to assume various restrictions, but anyway they won the best paper award.
BTW, I attended at another meeting yesterday evening. It was a really really boring and fucking one. It was a waste of time. :(
But I had happy time about 3 hours today, and I can go to bed with happy feeling. :)
Today, there were 4 speakers, and three of them gave us reports of SOSP '09, and the other one talked about Japan Linux Symposium.
For me, the most interesting topic was about formal verification work of L4 microkernel(!?), seL4. Although the authors from NICTA, Australia, had to assume various restrictions, but anyway they won the best paper award.
BTW, I attended at another meeting yesterday evening. It was a really really boring and fucking one. It was a waste of time. :(
But I had happy time about 3 hours today, and I can go to bed with happy feeling. :)
Wednesday, November 04, 2009
Beyond the Doubting of a Shadow

Beyond the Doubting of a Shadow
Roger Penrose (commentaries by Ken-ichiro Mogi and Kaoru Takeuchi)
ISBN 4-480-09006-1 (Japanese edition)
This book is not just a translation of Penrose's articles, but a collection of commentaries and discussions about Penrose's ideas on consciousness. Thus, Japanese title of the book is "ペンローズの量子脳理論" which means "The Quantum Brain Theory of Penrose."
Readers are required to have knowledge on not only Cognitive Science but also Computability, Mathematical Logic including Goedel's theory and contemporary Physics or background to understand them at least.
But, the translators, Mogi and Takeuchi, have wide variety of knowledge covering them, and their additional writings help us greatly understanding the original Penrose's idea. In this sense, the book is a unique work on the theme.
BTW, at least for me, I'm feeling that I need to read some more books to have enough understanding on the concept :(, and I'm looking fora good book on Mathematical Logic now....
Sunday, September 06, 2009
Season of Decay

Recently, my wife's SONY VAIO type F failed, and today I was trying to salvage data from its HDD. The disk was not broken, and I could successfully pick up necessary files which my wife needs. It's OK so far.
But, it turned out that CRT monitor was also broken when I tried to connect the HDD to my desktop PC. :(
Actually, it I bought it about 15 years ago, and maybe the CRT monitor, Nanao (EIZO) 53T proved its high reliability & quality...
Anyway, it looks like season of decay for me and take me costs... :(
Subscribe to:
Posts (Atom)



