WHO IS 2HEI?

ethtool maybe caused by "TCP checksum offload"

Wed, 03 Mar 2010 15:14:10 +0800

使用tcpdump抓包时发现有这样的错误信息：

Transmission Control Protocol, Src Port: 44937 (44937), Dst Port: https (443), Seq: 111, Ack: 147, Len: 6

Checksum: 0x5edd [incorrect, should be 0x15db (maybe caused by "TCP checksum offload"?)]
Good Checksum: False
Bad Checksum: True

或许是TCP checksum offload的原因

查看本机的网卡信息
2hei# ethtool -k eth0
Offload parameters for eth0:
Cannot get device rx csum settings: Operation not supported
Cannot get device udp large send offload settings: Operation not supported
rx-checksumming: off
tx-checksumming: on
scatter-gather: on
tcp segmentation offload: off
udp fragmentation offload: off
generic segmentation offload: off

关闭rx和tx
#ethtool -K rx off
#ethtool -K tx off
#ethtool -K eth0 rx off tx off tso off gso off

也可用这个命令
#ethtool -K eth0 tx off tso off

打开
#ethtool -K eth0 tx on
#ethtool -K eth0 tso on

关闭tx和rx后，再次抓包，结果显示正常。

CheckSum Offload实际上是将传输层的一部分工作交给了硬件完成，以节约系统的CPU资源。wireshark在本地抓包时看到是系统随机填充的校验和，所以会显示Checksum Offload，实际并不影响数据包的正常传输。
微软的测试表明它可以最多节约30%的CPU资源。IBM里AIX的文档则指出：对于PCI接口的千兆网卡来说还不如让400Mhz以上的CPU来计算校验和，而PCI-X的千兆网卡启用此项后可以达到线路速度，从而节约CPU资源。

详细可见： http://www.microsoft.com/whdc/device/network/taskoffload.mspx

request.getScheme cann't get https in nginx proxy with apache

Sun, 21 Feb 2010 15:37:43 +0800

程序中为了不写死url，使用了动态获得的方式：
basePath = request.getScheme()+"://"+request.getServerName()+":"+request.getServerPort()+path+"/";

其中： request.getScheme() return http but not https.
之前单独使用apache（https）+resin的方式正常，现在前面增加了一层nginx，发现问题来了，协议部分（Scheme）无法传过去，后台的resin无法获取到正确的值。

尝试了下面的配置，结果还是一无所获。
    proxy_redirect off;
    proxy_set_header HTTPS on;
    proxy_set_header Host $host;
    proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
    proxy_set_header X-Url-Scheme $scheme;
    proxy_set_header X-Scheme $scheme;
    proxy_set_header X-Nginx-Scheme $scheme;
    proxy_set_header Scheme $scheme;
    proxy_set_header X-FORWARDED_PROTO $scheme;
    proxy_set_header X-FORWARDED_PROTO "https";
    proxy_set_header X-Forwarded-Scheme "https";
    proxy_set_header X-Forwarded-Proto $scheme;
    real-scheme-header     X-Forwarded-Proto;

今天下午好好google了一番，所有的帖子几乎翻了个遍，终于有所收获啦，重点是nginx配置完毕后，apache也要配置环境变量哦！
http://www.ruby-forum.com/topic/183450

nginx config:
=============
proxy_set_header X-Nginx-Scheme $scheme;
# nginx variable $scheme will be 'http' or 'https'.

apache config:
==============
SetEnvIf X-Nginx-Scheme "^https$" HTTPS=on
# Apache environment variable HTTPS will be 'on' or not defined.

测试代码如下：
<%
out.println("Protocol: " + request.getProtocol() + "
");
out.println("Scheme: " + request.getScheme() + "
");
out.println("Server Name: " + request.getServerName() + "
" );
out.println("Server Port: " + request.getServerPort() + "
");
%>

页面打印结果为，正常了！
Protocol: HTTP/1.0
Scheme: https
Server Name: 2hei.net
Server Port: 443

resin使用了struts/spring等框架与apache整合时需要注意事项

Sat, 06 Feb 2010 12:09:44 +0800

resin使用了struts/spring等框架与apache整合时需要注意以下：

默认情况resin ResinConfigServer配置不支持除了jsp和servlet之外的其他不规则映射，apache对此类URL会返回404，或者无效的servlet，这就用到了resin的plugin_match配置。
如：

几种常见的resin解析方式：

#jsp文件

#标准的servlet

#resin自定义的一种格式

#扩展匹配，如struts

device lo left promiscuous mode

Thu, 28 Jan 2010 14:40:32 +0800

dmesg发现有一些这样的信息：

device lo left promiscuous mode
audit(1264387352.387:24): dev=lo prom=0 old_prom=256 auid=4294967295 ses=4294967295

发现是tcpdump 抓包导致,对系统并无大碍！

《网瘾战争》--看你妹

Sun, 24 Jan 2010 13:27:47 +0800

一口气看完了60多分钟的《网瘾战争》，真是有些感动。。。

我和其他热爱这个游戏的人一样
认真地挤着公交车上班
认真地消费着各式各样的食品
不管里面有没有不认识的化学成分
我们没有因为工资微薄而抱怨过
没有因为你们拿着从我微薄工资里扣的税
住着联体排屋而心理失衡过
在这一年里，我和其他热爱这个游戏的人一样
为水灾，为地震而痛心哭泣
为载人航天，为奥运会而加油喝彩
我们打心眼里，就不想我们在任何一个方面
落后于这个世界上的任何一个民族
而这一年里，却因为你们这些人
我们迟迟不能与地球上其他国家的玩家一起，一决高下
为了真心喜爱的游戏
我们委曲求全，我们被迫离开
我们冒着被封号的危险去美服欧服
他们骂我们是金币农民
我们顶着上万的延时去台服
他们说我们是大陆蝗虫
这些何其贬低的称谓
我们都在默默地承受
我们凭什么不能拥有每小时4毛钱的廉价娱乐？
就凭我在国服？
凭我是国服的魔兽玩家吗？
你们从小就对我灌输
金窝、银窝不如自己的狗窝
那现实呢？
你们已经让我只能暂住在自己的国家
难道我们精神上的家园
连暂住在自己的国家，都不行吗？

ps: youku视频已被河蟹！
土豆上的视频不知道还能撑多久？

linux下支持https的压力测试工具

Sun, 24 Jan 2010 11:02:42 +0800

测试了linux下的几种压力测试工具，发现有些不支持https，先简单总结如下：

一、apache的ab工具

/home/webadm/bin/ab -c 50 -n 10000 https://2hei.net/mt/index.html
SSL not compiled in; no https support
看样子是说SSL没有编译进来，所以不支持https

二、apache的flood工具

http://httpd.apache.org/test/flood/

wget http://www.apache.org/dist/httpd/flood/flood-0.4.tar.gz

如果要支持https的话，需要添加如下的编译参数：--with-apr --with-apr-util --enable-ssl
具体代码也可以从svn获取。
不过我在编译的时候遇到一些问题，apr-util包在make的时候总是报错！

三、web-bench工具

http://freshmeat.net/projects/web-bench/

./webbench -c 20 -t 10 https://2hei.net/mt/index.html
Webbench - Simple Web Benchmark 1.5
Copyright (c) Radim Kolar 1997-2004, GPL Open Source Software.

Only HTTP protocol is directly supported, set --proxy for others.
明确提示了提示不支持https哦！

四、http_load工具，曾经的最爱，http测试的结果还很令人满意。

http://acme.com/software/http_load/
./http_load -rate 5 -seconds 10 urls
./http_load: unknown protocol - https://2hei.net/mt/index.html
吼吼，看来之前常用的http_load也不支持https

五、siege工具

http://www.joedog.org/index/siege-home

编译使之支持https
./configure --prefix=/home/2hei.net/siege --with-ssl=/usr/include/openssl

基本用法：
./siege
SIEGE 2.69
Usage: siege [options]
       siege [options] URL
       siege -g URL
Options:
-V, --version           VERSION, prints version number to screen.
-h, --help              HELP, prints this section.
-C, --config            CONFIGURATION, show the current configuration.
-v, --verbose           VERBOSE, prints notification to screen.
-g, --get               GET, pull down headers from the server and display HTTP
                          transaction. Great for web application debugging.
-c, --concurrent=NUM    CONCURRENT users, default is 10
-u, --url="URL"         Deprecated. Set URL as the last argument.
-i, --internet          INTERNET user simulation, hits the URLs randomly.
-b, --benchmark         BENCHMARK, signifies no delay for time testing.
-t, --time=NUMm         TIME based testing where "m" is the modifier S, M, or H
                          no space between NUM and "m", ex: --time=1H, one hour test.
-r, --reps=NUM          REPS, number of times to run the test, default is 25
-f, --file=FILE         FILE, change the configuration file to file.
-R, --rc=FILE           RC, change the siegerc file to file. Overrides
                          the SIEGERC environmental variable.
-l, --log               LOG, logs the transaction to PREFIX/var/siege.log
-m, --mark="text"       MARK, mark the log file with a string separator.
-d, --delay=NUM         Time DELAY, random delay between 1 and num designed
                          to simulate human activity. Default value is 3
-H, --header="text"     Add a header to request (can be many)
-A, --user-agent="text" Sets User-Agent in request

siege -c 20 -r 2 -f url
-c 20 并发20个用户
-r 2 重复循环2次
-f url 任务列表：URL列表

结论相当凑合，纵欲找到可以支持https压力测试的工具了。

六、httperf工具，来自hp的工具，不过已经n年没有更新了。

wget ftp://ftp.hpl.hp.com/pub/httperf/httperf-0.9.0.tar.gz

        $ mkdir build
        $ cd build
        $ SRCDIR/configure
        $ make
        $ make install

工具使用方法：
/usr/local/bin/httperf --help
Usage: httperf [-hdvV] [--add-header S] [--burst-length N] [--client N/N]
        [--close-with-reset] [--debug N] [--failure-status N]
        [--help] [--hog] [--http-version S] [--max-connections N]
        [--max-piped-calls N] [--method S] [--no-host-hdr]
        [--num-calls N] [--num-conns N] [--period [d|u|e]T1[,T2]]
        [--port N] [--print-reply [header|body]] [--print-request [header|body]]
        [--rate X] [--recv-buffer N] [--retry-on-failure] [--send-buffer N]
        [--server S] [--server-name S] [--session-cookies]
        [--ssl] [--ssl-ciphers L] [--ssl-no-reuse]
        [--think-timeout X] [--timeout X] [--uri S] [--verbose] [--version]
        [--wlog y|n,file] [--wsess N,N,X] [--wsesslog N,X,file]
        [--wset N,X]
如：
/usr/local/bin/httperf --server www.2hei.net \
   --port 443 --uri /mt/index.html \
   --rate 15 --num-conn 1000 \
   --num-call 1 --timeout 5

发现会有如下错误：
httperf: warning: open file limit > FD_SETSIZE; limiting max. # of open files to FD_SETSIZE
以下是解决办法：
# Edit /etc/security/limits.conf and add the line * hard nofile 65535 (or instead of * you can put the username of the user for whom you want to change the limit)
# Edit /usr/include/bits/typesizes.h and change #define __FD_SET_SIZE 1024 to #define __FD_SET_SIZE 65535 (in /usr/include/sys/select.h FD_SETSIZE is defined as __FD_SETSIZE)
重新编译即可。

实例测试如下：
/usr/local/bin/httperf --client=0/1 --ssl --server=www.2hei.net --port=443 --uri=/mt/index.html --rate=1 --num-conns=1000 --rate=50 --num-calls=1 --hog
httperf --hog --client=0/1 --server=www.2hei.net --port=443 --uri=/mt/index.html --rate=50 --send-buffer=4096 --recv-buffer=16384 --ssl --num-conns=1000 --num-calls=1
Maximum connect burst length: 1

Total: connections 1000 requests 1000 replies 1000 test-duration 33.758 s

Connection rate: 29.6 conn/s (33.8 ms/conn, <=394 concurrent connections)
Connection time [ms]: min 73.1 avg 6513.5 max 22013.1 median 5371.5 stddev 4176.3
Connection time [ms]: connect 5670.7
Connection length [replies/conn]: 1.000

Request rate: 29.6 req/s (33.8 ms/req)
Request size [B]: 80.0

Reply rate [replies/s]: min 26.6 avg 31.3 max 33.6 stddev 2.4 (6 samples)
Reply time [ms]: response 842.7 transfer 0.0
Reply size [B]: header 331.0 content 163.0 footer 2.0 (total 496.0)
Reply status: 1xx=0 2xx=1000 3xx=0 4xx=0 5xx=0

CPU time [s]: user 9.91 system 23.11 (user 29.4% system 68.5% total 97.8%)
Net I/O: 16.6 KB/s (0.1*10^6 bps)

Errors: total 0 client-timo 0 socket-timo 0 connrefused 0 connreset 0
Errors: fd-unavail 0 addrunavail 0 ftab-full 0 other 0

nginx authentication

Wed, 20 Jan 2010 21:11:55 +0800

今天配置nginx的basic authentication折腾了好半天，开始error_log一直报错：
no user/password was provided for basic authentication

我是按照nginx的wiki配置的：
location / {
auth_basic "Restricted";
auth_basic_user_file htpasswd;
}

cat htpasswd
2hei:j3M4coizxFLDM

Since version 0.6.7 the filename path is relative to directory of nginx configuration file nginx.conf, but not to nginx prefix directory.

google了半天，发现了nginx作者Igor的一个回复：

Igor Sysoev
The HTTP Basic authentication works as following:
*) A browser requests a page without user/password.
*) A server response with 401 page, sending realm as well.
   At this stage the 401 code appears in access_log and the message
   "no user/password ..." appears in error_log.
*) The browser shows a realm/login/password prompt.
*) If a user will press cancel, then the browser will show the received
   401 page.
*) If the user enters login/password, then the browser repeats the request
   with login/password.

Then until you will exit the browser, it will send these login/password
with all requests in protected hierarchy.

赞一个，牛人果然就是牛人！

经过提示找到了罪魁祸首是error_page的配置 401 403 404 /40x.html;

error_page 401 403 404 /40x.html;
实际上是由于40x.html文件并不存在导致的。把文件建好问题得以解决！

[notice] child pid 19818 exit signal Bus error (7)

Thu, 07 Jan 2010 14:47:45 +0800

apache logs:
[Thu Dec 31 12:27:41 2009] [notice] child pid 19818 exit signal Bus error (7)

The first error "Bus Error" is caused by someones (possibly CGI) program crashing. A bus error happens when someone wrote a C program for Intel and re-compiled it for another processor like PPC or Mips that does not support misalligned memory access. Really all it means is someone is running buggy software on your server machine.
The second error could also be a CGI program that hangs longer than the server's timeout, so its killed. Probably another buggy program.

That sounds like bad RAM or maybe a bad chassis. I would try and build something other the Apache and see if you get similar errors. If they persist then I would ask for a RAM swap to be done and check again. If that does not fix it then request a chassis swap.

There is a small chance this could be OS related if you are running a different Kernel release compared to the OS default or have many patches installed on it but I am leaning towards a hardware issue.

http://www.kernel.org/doc/man-pages/online/pages/man7/signal.7.html

Standard Signals

       Linux supports the standard signals listed below. Several signal numbers are
       architecture-dependent, as indicated in the "Value" column. (Where three
       values are given, the first one is usually valid for alpha and sparc, the
       middle one for ix86, ia64, ppc, s390, arm and sh, and the last one for mips.
       A - denotes that a signal is absent on the corresponding architecture.)

       First the signals described in the original POSIX.1-1990 standard.

       Signal     Value     Action   Comment

       ----------------------------------------------------------------------
       SIGHUP        1       Term    Hangup detected on controlling terminal
                                     or death of controlling process
       SIGINT        2       Term    Interrupt from keyboard
       SIGQUIT       3       Core    Quit from keyboard
       SIGILL        4       Core    Illegal Instruction
       SIGABRT       6       Core    Abort signal from abort(3)
       SIGFPE        8       Core    Floating point exception
       SIGKILL       9       Term    Kill signal
       SIGSEGV      11       Core    Invalid memory reference
       SIGPIPE      13       Term    Broken pipe: write to pipe with no
                                     readers
       SIGALRM      14       Term    Timer signal from alarm(2)
       SIGTERM      15       Term    Termination signal
       SIGUSR1   30,10,16    Term    User-defined signal 1
       SIGUSR2   31,12,17    Term    User-defined signal 2
       SIGCHLD   20,17,18    Ign     Child stopped or terminated
       SIGCONT   19,18,25    Cont    Continue if stopped
       SIGSTOP   17,19,23    Stop    Stop process
       SIGTSTP   18,20,24    Stop    Stop typed at tty
       SIGTTIN   21,21,26    Stop    tty input for background process
       SIGTTOU   22,22,27    Stop    tty output for background process

       The signals SIGKILL and SIGSTOP cannot be caught, blocked, or ignored.

       Next the signals not in the POSIX.1-1990 standard but described in SUSv2 and
       POSIX.1-2001.

       Signal       Value     Action   Comment
       --------------------------------------------------------------------
       SIGBUS      10,7,10     Core    Bus error (bad memory access)
       SIGPOLL                 Term    Pollable event (Sys V).
                                       Synonym for SIGIO
       SIGPROF     27,27,29    Term    Profiling timer expired
       SIGSYS      12,-,12     Core    Bad argument to routine (SVr4)
       SIGTRAP        5        Core    Trace/breakpoint trap
       SIGURG      16,23,21    Ign     Urgent condition on socket (4.2BSD)
       SIGVTALRM   26,26,28    Term    Virtual alarm clock (4.2BSD)
       SIGXCPU     24,24,30    Core    CPU time limit exceeded (4.2BSD)
       SIGXFSZ     25,25,31    Core    File size limit exceeded (4.2BSD)

使用apache的mod_rewrite模块进行端口重写且保留特殊接口

Thu, 07 Jan 2010 14:35:11 +0800

需求如下：
网站的apache开通80和443端口，整站强制使用https，把所有80端口的访问转到443，新增需求为要针对某个接口单独开放80端口（80、443均可访问），剩下其他接口保持原有规则。

配置一、
RewriteRule !^/TestServlet http://%{SERVER_NAME}/$1 [L,R=301] 是有问题的，$1并没取得匹配的部分。

根据apache文档，使用否字符以反转匹配模式时，匹配模式中不能使用分组的通配成分。由于模式不匹配而使分组的内容是空的，所以它是不可能实现的。因此，如果使用了否定式匹配模式，那么后继的字符串中就不能使用$N。

配置二、
RewriteRule !^/TestServlet - [L,R=301]   生效。

apache文档说明：'-'是一个特殊的替换串，意思是不替换。它可以用于仅仅匹配某些URL而无须替换的情况，比如，在发生替换前，允许以C(chain)标记连接的多个匹配模式同时起作用。

配置三、
    RewriteRule ^(.*)?$ https://%{SERVER_NAME}$1 [L,R=301]   原有将http重写为https的配置。
    所有访问http://2hei.net会转到https://2hei.net

另外或许可行办法是配置apache的系统变量或者环境变量，凡是经过url重写的给变量一个特殊的标记，说明已经跳转，这样就不会出现循环跳转的现象了（没具体试过！）

实际中遇到的一个问题是apache进行url重写时，post方式提交的数据经过重写后数据会丢失，跳转后的url并没有收到post的数据，而访问跳转后的页面成了get方式。暴汗！

网上有人说使用 P参数，
RewriteRule ^(.*)?$ https://%{SERVER_NAME}$1 [L,R=301,P] ，实际试用了一下，纯属扯淡！

'proxy|P'(强制为代理)
此标记使替换成分被内部地强制作为代理请求发送，并立即中断重写处理，然后把处理移交给mod_proxy模块。你必须确保此替换串是一个能够被mod_proxy处理的有效URI(比如以http://hostname开头)，否则将得到一个代理模块返回的错误。使用这个标记，可以把某些远程成分映射到本地服务器域名空间，从而增强了ProxyPass指令的功能。
注意：要使用这个功能，必须已经启用了mod_proxy模块。

8.8.8.8-google-public-dns-for-free

Wed, 30 Dec 2009 09:44:41 +0800

http://code.google.com/speed/public-dns/index.html

What is Google Public DNS?

Google Public DNS is a free, global Domain Name System (DNS) resolution service, that you can use as an alternative to your current DNS provider.

To try it out:

* Configure your network settings to use the IP addresses 8.8.8.8 and 8.8.4.4 as your DNS servers or
* Read our configuration instructions.

If you decide to try Google Public DNS, your client programs will perform all DNS lookups using Google Public DNS.

实际测试8.8.8.8 解析出来的貌似网通的IP，不过I like it，好记又方便，而且无dns劫持的弊端，不错！唯一担心的是有可能被墙哦！

nagios check_http 处理返回码302的情况

Wed, 16 Dec 2009 16:44:58 +0800

nagios check_http 处理返回码302的情况

使用nagios监控中发现这样一个问题，nagios的check_http对http返回码302的处理比较特殊，如果参数不当的话会影响监控的效果哦！

实际情况如下：
在web服务器配置的errpage页面指定了503跳转，如apache+resin的组合，如果apache找不到resin会返回503，捕获的页面会跳转到指定的页面上

1、正常情况：
/home/nagios/libexec/check_http -S -H 2hei.net -I 10.10.10.10 -u "/monitor.jsp" -P "" -s "OK" -t 10
返回
HTTP OK HTTP/1.1 200 OK - 617 bytes in 0.345 seconds |time=0.345145s;;;0.000000 size=617B;;;0

如果页面返回值不包含“OK”，则
HTTP CRITICAL - string not found|time=0.034220s;;;0.000000 size=617B;;;0

如果停掉保留apache不动，停掉resin，则以外发生了，302跳转后状态仍然是OK，I服了U ！！！
HTTP OK - HTTP/1.1 302 Found - 0.030 second response time |time=0.029908s;;;0.000000 size=517B;;;0

加入-v 参数后看看详细结果：
https://2hei.net:443/monitor.jsp is 517 characters
STATUS: HTTP/1.1 302 Found
**** HEADER ****
Date: Wed, 16 Dec 2009 07:49:53 GMT
Server: Apache
Location: https://2hei.net/errorcode/503.html
Cache-Control: max-age=0
Expires: Wed, 16 Dec 2009 07:49:53 GMT
Vary: Accept-Encoding
Content-Length: 220
Connection: close
Content-Type: text/html; charset=iso-8859-1
**** CONTENT ****

302 Found

Found

The document has moved here.

HTTP OK - HTTP/1.1 302 Found - 0.030 second response time |time=0.029908s;;;0.000000 size=517B;;;0

丫难道是nagios check_http的bug？

在仔细看看nagios check_http 用法：

Usage: check_http -H | -I [-u ] [-p ]
       [-w ] [-c ] [-t ] [-L]
       [-a auth] [-f ] [-e ]
       [-s string] [-l] [-r | -R ] [-P string]
       [-m :] [-4|-6] [-N] [-M ] [-A string]
       [-k string] [-S] [-C ] [-T ]

常用参数
-H 用于虚拟主机，并可以增加端口。例如，test.com:10000
-I IP地址
-S 基于SSL连接。默认端口为443
-u URL地址 “/”
-w 警告响应时间响应时间超过指定时间则显示WARNING
-c 临界报警响应时间响应时间超过设定时间则显示CRITICAL
-p 端口设定
-e 监测服务器反馈的第一行(状态)信息是否包含匹配字符
-N 不读取body部分，在读取header信息，这样相应速度比较快
-t 超时时间值
-v 显示服务器反馈的所有信息
-r 正则表达式匹配反馈页信息
-R 同上，但可以不考虑大小写问题
--invert-regex 反馈信息中包含字符则报警
-L 反馈信息包含URL link
-a 对需要基础认证的，提供用户名和密码，获取反馈值

尝试使用了-e参数，问题搞定！ yeah ！！！

-e 指定check_http返回的第一行status中包含的字符串
STATUS: HTTP/1.1 302 Found

/home/nagios/libexec/check_http -S -H 2hei.net -I 10.10.10.10 -u "/monitor.jsp" -P "" -s "OK" -t 10 -e "200"
这样只有返回码包含200的才会OK，其他状态码一律FATAL！
最后的command如下：
define command{
        command_name    check_https
        command_line    $USER1$/check_http -S -H $ARG1$ -I $HOSTNAME$ -u $ARG2$ -P $ARG3$ -s $ARG4$ -t $ARG5$ -e $ARG6$
        }

nagios的check_http果然很牛X。

xenconsole: Could not read tty from store: No such file or directory

Fri, 27 Nov 2009 15:47:27 +0800

#xem在配置heartbeat时遇到的一个问题，重启虚拟机出现了问题，系统无法启动，控制台也无法登陆。
xm console xm-2hei.net
xenconsole: Could not read tty from store: No such file or directory

原来是xenconsoled进程不在了，重启即可
#/usr/sbin/xenconsoled

另外一个xen的小tip是，“]”键可以方便的从控制台退出来，哈哈！

Read-only file system on CentOS5

Wed, 25 Nov 2009 11:54:56 +0800

We have 5 Dell server collocated running CentOS 4.x and 5.x server operating system. Sometime my file system (ext3) goes read-only.
I’d like to know what could be causing such a problem?

CentOS 5.x Read-only file system

form http://www.cyberciti.biz/tips/linux-filesytem-goes-read-only.html:
a) Hardware problem / hard disk problem, check harddisk for errors.

b) High disk I/O aka busy I/O retry error can mark low level disk call as failed. This will force ext3 to go into read only mode.

c) High disk I/O on SAN

d) SAN is not configured properly for the path failover.

In all sort of problems ext3 goes read-only to protect the filesystem and further damage. If you are using VMWARE, check out official webpage to download SCSI patches or workaround for vmware problems.
So what could be causing the file system on Linux go read-only?

Apart from above generic problem, any other error can trigger filesystem on Linux go read only. I hope our reader / seasoned Linux admin can help to answer this question.

reboot my OS , file system recovery.

and you can try others:
mount -o remount,rw /
mount -o remount /

mptscsih: ioc0: attempting task abort

Mon, 23 Nov 2009 13:43:13 +0800

Nov 19 00:03:14 2hei.net kernel: mptscsih: ioc0: attempting task abort! (sc=ffff81004f0c7380)
Nov 19 00:03:14 2hei.net kernel: sd 0:0:0:0:
Nov 19 00:03:14 2hei.net kernel: command: Read(10): 28 00 00 26 15 d7 00 00 48 00
Nov 19 00:03:14 2hei.net kernel: mptbase: ioc0: LogInfo(0x31140000): Originator={PL}, Code={IO Executed}, SubCode(0x0000)
Nov 19 00:03:14 2hei.net kernel: mptscsih: ioc0: task abort: SUCCESS (sc=ffff81004f0c7380)
Nov 19 00:03:24 2hei.net kernel: mptscsih: ioc0: attempting task abort! (sc=ffff81004f0c7380)
Nov 19 00:03:24 2hei.net kernel: sd 0:0:0:0:
Nov 19 00:03:24 2hei.net kernel: command: Test Unit Ready: 00 00 00 00 00 00
Nov 19 00:03:25 2hei.net kernel: mptbase: ioc0: LogInfo(0x31130000): Originator={PL}, Code={IO Not Yet Executed}, SubCode(0x0000)
Nov 19 00:03:25 2hei.net kernel: mptscsih: ioc0: task abort: SUCCESS (sc=ffff81004f0c7380)
Nov 19 00:03:25 2hei.net kernel: mptscsih: ioc0: attempting task abort! (sc=ffff81006dce49c0)
Nov 19 00:03:25 2hei.net kernel: sd 0:0:0:0:
Nov 19 00:03:25 2hei.net kernel: command: Write(10): 2a 00 04 c4 a5 f2 00 01 40 00
Nov 19 00:12:10 2hei.net kernel: mptbase: ioc0: LogInfo(0x31140000): Originator={PL}, Code={IO Executed}, SubCode(0x0000)
Nov 19 00:12:11 2hei.net kernel: mptscsih: ioc0: task abort: SUCCESS (sc=ffff81006dce49c0)
Nov 19 00:12:11 2hei.net kernel: mptscsih: ioc0: attempting task abort! (sc=ffff81006dce49c0)
Nov 19 00:12:11 2hei.net kernel: sd 0:0:0:0:
Nov 19 00:12:11 2hei.net kernel: command: Test Unit Ready: 00 00 00 00 00 00
Nov 19 00:12:11 2hei.net kernel: mptbase: ioc0: LogInfo(0x31130000): Originator={PL}, Code={IO Not Yet Executed}, SubCode(0x0000)
Nov 19 00:12:11 2hei.net kernel: mptscsih: ioc0: task abort: SUCCESS (sc=ffff81006dce49c0)
Nov 19 00:12:11 2hei.net kernel: mptscsih: ioc0: attempting task abort! (sc=ffff81007f0d4980)

Linux 2hei.net.163.com 2.6.18-53.1.13.el5 #1 SMP Tue Feb 12 13:02:30 EST 2008 x86_64 x86_64 x86_64 GNU/Linux

[root@2hei.net ~]# cat /proc/scsi/scsi
Attached devices:
Host: scsi0 Channel: 00 Id: 00 Lun: 00
Vendor: HP Model: DF072A9844 Rev: HPD0
Type: Direct-Access ANSI SCSI revision: 05

[root@2hei.net ~]# modinfo mptsas
filename: /lib/modules/2.6.18-53.1.13.el5/kernel/drivers/message/fusion/mptsas.ko
version: 3.04.04
license: GPL
description: Fusion MPT SAS Host driver
author: LSI Logic Corporation
srcversion: D0507BBC402F30BCC5539C0
alias: pci:v00001000d00000062sv*sd*bc*sc*i*
alias: pci:v00001000d00000058sv*sd*bc*sc*i*
alias: pci:v00001000d00000056sv*sd*bc*sc*i*
alias: pci:v00001000d00000054sv*sd*bc*sc*i*
alias: pci:v00001000d00000050sv*sd*bc*sc*i*
depends: mptscsih,mptbase,scsi_mod,scsi_transport_sas
vermagic: 2.6.18-53.1.13.el5 SMP mod_unload gcc-4.1
parm: mpt_pt_clear: Clear persistency table: enable=1 (default=MPTSCSIH_PT_CLEAR=0) (int)
parm: max_lun: max lun, default=16895 (int)

Do you have acpi=off parameter passed to kernel? Try to set it on.
Also, do you use ReiserFS as filesystem?

http://h20000.www2.hp.com/bizsupport/TechSupport/Document.jsp?objectID=c00715023&lang=en&cc=us&taskId=101&prodSeriesId=428936&prodTypeId=15351
http://h50176.www5.hp.com/support/292889-291/solve/108727.html

linux下用python实现文本menu

Tue, 17 Nov 2009 18:13:49 +0800

linux用shell实现menu比较简单，今天用python实现了一个，因为python没有switch语法，所以使用了dict来替代，另外调用了linux的clear来清屏，本例只是说明一下简单的实现方法，菜单的命令或者语句可以自由发挥呵呵。

截个图先：

具体代码如下：
#!/usr/bin/evn python
# -*- coding: utf-8 -*-
#Author: 2hei#2hei.net
#Date: 2009-11-17 18:24

import os,sys

running = True
menu = """
             menu
------------------------------
    1:   Disk info
    2:   Mem info
    3:   Network info
    4:   Sys load info
    5:   Process info
    h:   Help
    q:   Quit
------------------------------
"""

menu_dict={
      "h": "echo help ^_^",
      "1": "df -h",
      "2": "free -m",
      "3": "netstat -lnt",
      "4": "uptime",
      "5": "ps x"
      }

def commands(args):
    cmd = menu_dict.get(args)
    return cmd

if __name__ == "__main__":
    os.system('clear')
    print menu
    while running:
        cmd = raw_input("Input your commond :\n")
        if cmd != 'q':
            os.system('clear')
            try:
                print menu
                if commands(cmd) != None:
                    fo = os.popen(commands(cmd))
                    print fo.read()
                else:
                    print "Input is Wrong!\n"
            except Exception,e:
                print menu
                print e
        else:
            print 'we will exit the menu\n'
            sys.exit()