= 2010-06-21 =

== Linux ==

 * [http://www.liamdelahunty.com/tips/linux_ls_awk_totals.php Linux - Sum using awk] - 用 awk 指令進行檔案大小加總(sum)
{{{

$ ls -lh php*; ls -l php* | awk '{ SUM += $5} END { print SUM/1024/1024 }'

-rw-rw-r--    1 admin    site13       1013 Aug 12 11:01 php_checkbox_check.php
-rw-rw-r--    1 admin    site13       8.9k Nov 11 23:15 php_contact_form.php
-rw-rw-r--    1 liamdela site13        560 Oct  1  2002 php_convert_url_to_link.php
-rw-r--r--    1 root     site13        948 Dec  7 21:56 php_gethostbyaddr.php
-rw-rw-r--    1 liamdela site13       2.1k Feb 25  2002 php_list_a_directory.php
-rw-rw-r--    1 liamdela site13       1.7k May 28  2003 php_multi_dimensional_arrays.php
-rw-rw-r--    1 admin    site13       2.9k Nov 14 13:15 php_prev_next_array.php
-rw-r--r--    1 admin    site13       1.9k Feb  3 00:55 php_uk_mailing_list.php

0.0194302
}}}

 * [http://ubuntuforums.org/showthread.php?t=424366 rsync resume downloads] - 使用 rsync 作檔案續傳動作
{{{
$ rsync --partial --progress --rsh=ssh localfile remotehost:directory/
}}}

== Hadoop ==

 * 今天在論壇上有人問到如果 google 的查詢索引放在記憶體中，若使用 Hadoop 與 HBase 的邏輯來思考，資料流大概是長這樣：
{{{
#!graphviz
digraph cloud_stack { size="16,10"; node[shape=box,width=2.0];
   "Crawlers" -> "HTML files stored in Google File System" -> "Map/Reduce Job to analysis keywords into index" -> "(search keyword, URLs) key-value pairs stored in BigTable";
}
}}}