Rails Installation guide and quick fix tips

Redis & nginx configuration for websocket/actionCable

2019-06-19T00:16:00.000-07:00

Start redis server => redis-server
redis console => redis-cli

view available databases => config get databases

pubsub channels global*
pubsub channels global_notification_2
pubsub channels global_notification

pubsub channels * => Lists all the available channels
pubsub numsub global_notification_2 => returns the number of subscribers

From unix command line directly

redis-cli PUBSUB CHANNELS
redis-cli PUBSUB NUMSUB global_notification_2

[root@api prod-openmednet-api]# redis-cli PUBSUB CHANNELS
1) "_action_cable_internal"
2) "global_notification_2"

[root@api prod-openmednet-api]# redis-cli PUBSUB CHANNELS global_notification_2
1) "global_notification_2"

[root@api prod-openmednet-api]# redis-cli PUBSUB NUMSUB global_notification_2

1) "global_notification_2"

2) (integer) 2

NGINX configuration:

worker_processes auto;

worker_rlimit_nofile 500000; #=> refer RLIMIT_NOFILE

server {

error_log logs/dev/error.log warn;

listen 80 default_server;

server_name devapi.openmednet.org;

root /home/rubyapps/OMNR5/OpenMedNet-R5/dev-openmednet-api/public;

rails_env development;

passenger_enabled on;

if ($http_user_agent ~* LWP::Simple|BBBike|jorgee) {

return 403;

}

access_log logs/dev/host.dev.access.log main;

error_page 500 502 503 504 /50x.html;

location = /50x.html {

root html;

}

location /cable {

passenger_app_group_name omndevapi_action_cable;

passenger_force_max_concurrent_requests_per_process 0;

}

setting RLIMIT_NOFILE

step: 1

vi /etc/security/limits.conf

add the following two lines at the end of the file

* soft nofile 120000

* hard nofile 120000

ref: https://caradede.blogspot.com/2017/12/how-to-solve-centos-7-nofile-limit-nginx.html

mkdir /etc/systemd/system/nginx.service.d

vi /etc/systemd/system/nginx.service.d/nofile_limit.conf

have the following two lines in that file

[Service]

LimitNOFILE=500000

systemctl daemon-reload systemctl restart nginx.service

or simply set

setsebool -P httpd_setrlimit 1

And then have the following 3 lines of code @ /etc/sysctl.conf

net.core.somaxconn = 4096

fs.file-max = 120000

vm.overcommit_memory = 1

Then don't forget to restart the machine to apply the settings. AFter restart

check "ulimit -n" it should match fs.file-max above

this command too will match

cat /proc/$(cat /opt/nginx/logs/nginx.pid)/limits | grep open.files

Mac stuffs

2019-06-13T01:05:00.003-07:00

vi command	description
0	move to beginning of the current line
$	move to end of line
H	move to the top of the current window (high)
M	move to the middle of the current window (middle)
L	move to the bottom line of the current window (low)
1G	move to the first line of the file
20G	move to the 20th line of the file
G	move to the last line of the file

Postgres points

2019-06-05T23:15:00.003-07:00

Export sql data to file

copy (select * from table-name) to '/tmp/cta.pl.list'

in root
dnf install postgresql-server postgresql-devel
service postgresql start

mkdir /home/pgdir
chown postgres:postgres /home/pgdir -R
chmod 777 /home/pgdir -R

su postgres
initdb -D /home/pgdir

pg_ctl -D /home/pgdir -l /home/pgdir/log start

Sublime

2019-05-09T04:11:00.002-07:00

1. Accessing remote files in sublime

ssh -R 52698:localhost:52698 root@192.xxx.xxx.xxx => run this command in local 

root@192.xxx.xxx.xxx  - it is the target remote ip from where you are referring the file to edit in local

Install remote sublime at remote server

curl -o /usr/local/bin/rmate https://raw.githubusercontent.com/aurora/rmate/master/rmate

sudo chmod +x /usr/local/bin/rmate

rmate filename

Linux/Unix useful commands

2019-05-09T04:09:00.000-07:00

Find count of files in a folder

ls -l /san1/home/ports/Quest/QuestHL7incoming/new/prod/comite_part1_2xx/ | wc -l

Enable/Disable firewall

CentOS V < 7

service iptables list

service iptables start

service iptables stop

CentOS V >= 7

Status: systemctl status firewalld 

Start: systemctl start firewalld 

Stop: systemctl stop firewalld 

Login to a server without password

run ssh-keygen from source machine will create key stored at ~/.ssh/id_rsa.pub copy the code alone and paste in the destination server > autorised_keys

/root/.ssh/authorized_keys => if not exists in the destination server, you can create it

Rails 5 - assets precompile issue - Easy way to trace

2019-05-09T04:06:00.000-07:00

Run the following command under rails console to trace which file raising issue upon precompiling

JS_PATH = "app/assets/javascripts/**/*.js";

Dir[JS_PATH].each do |file_name|
puts "\n#{file_name}"
puts Uglifier.compile(File.read(file_name))
end

UNIX - Process/Jobs

2015-05-17T22:03:00.002-07:00

Playing the commands fg , ctrl+z , bg and jobs

In this post I would like to share few basic process or job specific tricks in UNIX.

1. If you want to run a process in bacground just add a & as suffix with a space

Eg:

[root@localhost kannan]# ruby test.rb &

here this process test.rb will run in background

2. If you want list available background tasks/processes/jobs , just use the command 'jobs'

Eg:

[root@localhost kannan]# jobs

here the result would be

[1]+ Running ruby test.rb

3. If you want to bring a background process in front (foreground) just use fg [#<job number>]

[root@localhost kannan]# fg #1

It will bring the above 'ruby test.rb' process in front.

4. If you want to again move a already running process to background , use two combination of commands

a. First pass Ctrl+Z , here you will get a message like 'stopped' don't get upset, it will just suspend

b. Do run 'bg' command will take the suspended process/job/task to background process

[root@localhost kannan]# jobs

[1]+ Running ruby test.rb &

[root@localhost kannan]# fg #1

ruby test.rb

[1]+ Stopped ruby test.rb

[root@localhost kannan]# bg

[1]+ ruby test.rb &

[root@localhost kannan]#

RVM with rails1.2.1 and rails 3.2.x

2013-01-07T22:49:00.000-08:00

rvm install ext-rvm-list
\curl -L https://get.rvm.io | bash -s stable
source /etc/profile.d/rvm.sh
rvm automount
rvm list
whereis ruby
rvm list ruby
rvm list rubies
rvm install 1.8.7
rvm install 1.9.3
rvm list
rvm set default ruby-1.8.7-p371
rvm --help
rvm default ruby-1.8.7-p371
rvm list
gem list
yum install gcc-c++
yum install openssl
rvm pkg install openssl
rvm pkg install iconv
rvm list
rvm use ruby-1.9.3-p327
svn co http://b3.flexis.net:3001/omie_dev_r3
cd omie_dev_r3/
ls
rvm list
bundle install
yum install postgresql postgresql-devel
gem install pg -v '0.14.1'
bundle install
cd ..
cd omie_dev
ls
rvm list
rvm use ruby-1.8.7-p371
clear
ruby -v
gem install rails -v 1.2.1
gem install imagesize -v 0.1.1
gem install parseexcel 0.5.2
gem install payment -v 1.0.1
gem install pdf-writer -v 1.1.3
gem install pdfkit -v 0.5.0
gem install pg -v 0.14.0
gem install postgres -v 0.7.9.2008.01.28
gem install rmagick -v 2.13.1
gem install rubyzip -v 0.9.4
gem install simple_xlsx_writer -v 0.5.3
gem install spreadsheet -v 0.7.3
gem install tlsmail -v 0.0.1
gem install ruby-debug-ide -v 0.4.9
gem install ruby-ole -v 1.2.11.3
gem install ruby-debug-base -v 0.10.4

Errno::ENOENT: No such file or directory

2012-05-10T10:03:00.003-07:00

HTML to PDF conversion is very easy in rails and pdfkit will be useful in this case, wkhtmltopdf tool will be supporting tool so you can easily convert an html to pdf.

In my experience I used the following steps:

gem install pdfkit -v 0.5.0

gem install wkhtmltopdf -v 0.1.2

gem install wkhtmltopdf-binary

Some times this sequence will not work, so I use the following as per the instruction from https://github.com/pdfkit/pdfkit/wiki/Installing-WKHTMLTOPDF

Be careful to download the appropriate wkhtmltopdf tar ball, it depends on your os architecture, like i386 or 64

wget http://wkhtmltopdf.googlecode.com/files/wkhtmltopdf-0.9.9-static-amd64.tar.bz2
tar xvjf wkhtmltopdf-0.9.9-static-amd64.tar.bz2
mv wkhtmltopdf-amd64 /usr/local/bin/wkhtmltopdf
chmod +x /usr/local/bin/wkhtmltopdf

Try to convert html to pdf after restarting your web servers.

If you get any error like: Errno::ENOENT: No such file or directory then I have a straightforward method and it is very simple too.

Uninstall the gem "wkhtmltopdf"

gem uninstall wkhtmltopdf

then do the above things again, that is:

Restart your web servers, it should work.

Invalid json result

2012-05-09T05:37:00.000-07:00

In rails some times we are getting invalid json outputs say for example, as per the json standard, both the attributes and its values should be enclosed with double quotes (") but in some of the ruby patches has issues with that. If you are facing such issues then this post will be useful:

Changes what we have to make in the file 'core.rb' which is located at:

LIB_PATH/ruby/gems/1.8/gems/activesupport-1.4.0/lib/active_support/json/encoders

in your rails server, the LIB PATH is /usr/local/lib

required function is "encoder" line# 52

New Code

========

define_encoder Hash do |hash|

returning result = '{' do

result << hash.map do |key, value|

key = ActiveSupport::JSON::Variable.new(key.to_s) if

ActiveSupport::JSON.can_unquote_identifier?(key)

"\"#{key.to_json}\": #{value.to_json}" #=> This is line where we included the required quotes, thats it

end * ', '

result << '}'

end

Old Code

========

define_encoder Hash do |hash|

returning result = '{' do

result << hash.map do |key, value|

key = ActiveSupport::JSON::Variable.new(key.to_s) if

ActiveSupport::JSON.can_unquote_identifier?(key)

"#{key.to_json}: #{value.to_json}"

end * ', '

result << '}'

end

Output Sample: [Json attribute key should be enclosed with quotes]

Old code will produce the json output as follows:

[

{

person: {

id: 1,

last_name: "test"

}

]

New code will produce the json output as follows [Note the json attribute keys are enclosed with quotes]

[

{

"person": {

"id": 1,

"last_name": "test"

}

]

You can validate the json code at http://jsonlint.com/ , also the site http://jsonviewer.stack.hu/ will helps you understand the hierarchy of the json output.

openssl load error

2012-05-09T05:19:00.001-07:00

If you found any payment issue due to openssl (like require 'openssl' fails) then you have to be very careful.

Exact error might be:

LoadError: no such file to load -- openssl

This error occurred because you have installed your ruby without openssl support. So first install openssl then install ruby with openssl support.

If you still getting the error then you can follow the steps:

1. Check whether the openssl package is installed, if not please use 'yum install openssl'

2. Still the issue not solved then install openssl-dev, also make sure the library 'libruby' is installed

3. Still the issue not solved then check the ruby lib has the directory called 'openssl'

Note: In my server, the ruby lib is located at: /usr/local/lib/ruby/1.8/, better you can search the exact ruby's lib through 'whereis ruby' command

4. If openssl directory is not in your ruby lib then possibly you might get the openssl load error, then you can follow the simple steps:

a) copy openssl directory from any one of your other servers which is processing the 'openssl' and 'payment' things successfully

b) Then try 'require openssl' in your irb environment, If you get any "openssl.so" error, then search the file "openssl.so" and make a symbolic link to the x86_64-linux directory, if the file is not in your system then simply copy the openssl.so from other servers to 'x86_64-linux' that is /usr/local/lib/ruby/1.8/x86_64-linux'

5. Don't forget to restart your server (not only the web servers)

Going to convert your existing Rails 1.2.1 apps to 3.2.3 ? then this post may help you!!!

2012-05-04T10:11:00.001-07:00

Rails 1.2.1 => Rails 3.2.3

Actually, we started a product on 2007 with rails 1.2.1, at that time we were busy in the core development and we sincerely forgot to look about the rails updates!! in-between 3-4 years rails jumped to 3.x.x then only we realized that we are doing some funny things so started to review about 3.x.x

Almost we are out-dated, but still it is not too far .. We realized that this is right time to convert our existing 1.2.1 apps to 3.2.3 so first we have started to find the practical difficulties and possibilities and learned the followings. The following is the major things for us, you guys may meet some more, but any way it will be helpful for someone.

Lets checkout one by one.
1. RVM will helps us to manage different ruby and gems version in a same system

2. All required gems should be maintained through Gemfile, it will be useful when the time of deployment

3. You have to understand, learn and change the file route.rb

4. .rhtml is no longer in use, try to use either .html or .html.erb [I have created a script for this, let me know if any one wants] Note: .html.erb is the recommended extension

5. Need to use <%= yield %> instead of <%= @content_for_layout %> in layout

6. Here after all js, css and images are referred from ‘assets’ at /app directory instead of /public. and we can’t use /images/omie.gif here , need to use /assets/omie.gif in all of the places where we are referring images like <img> , css etc.,

7. Have to change pagination usage

@rss_pages, @rss_articles = paginate(:rss_articles, :order => "published_date DESC", :conditions => ["publication_string = ? ", "NEJM" ])

to

@rss_articles = RssArticle.paginate(:page => params[:page], :order => "published_date DESC", :conditions => ["publication_string = ? ", "NEJM" ])

Note: paginate will return only one object instead of two like what we have in our old (1.2.1) method, you can use the same object to get about page information, like @rss_articles.current_page, @rss_articules.next_page

In your old method you might uses, @rss_pages.current.next, here it is simple @rss_articles.next_page

8. Have to change paginate_by_sql

In your older method you might hacked the code in your application controller, but here the gem ‘will_paginate’ will do all of the required things for us. But of-course the usage of paginate_by_sql is bit different, look at our old method

@sql = “select * from lab_tests”

@unreviewed_pages,@labs_unreviewed = paginate_by_sql(LabTest, @sql,25)

but now we need to use as:

@labs_unreviewed = LabTest.paginate_by_sql(@sql,{:page => page, :per_page => 25})

9. use the function 'raw' to avoid HTML encoding
  <% a= “<div> Test </div>” %>
  <%= a %>
      The output will be a html encoded like <div< Test </div>
  <%= raw a %>
      The output like what you expected <div> Test </div>

10. All session values (> 4kb) must maintain only through active_record_store [We can setup it through config/initializers/session_store.rb] before it is done directly in environment.rb and this one is optional in lower version.
Note: Rails 3 uses a newer version of session schema, need to remove our old one and have to migrate the newer one.

11. We can’t use render(:layout=>false) in viewer, of-course it is meaningless but some of us used in some of our old view pages (like /comp/_datepicker.rhtml) by our old developers. :p

12. request.url is enough instead of request.request_uri

13. request.env in enough instead of @request.evn to get environment keys

14. We need special plugins to use ‘act_as’ modifiers such as acts_as_list, acts_as_tree.
it is under in review, as per my current knowledge , we have to use ‘ancestry’.
ref: https://github.com/stefankroes/ancestry

rails plugin install acts_as_tree => but it raises error. need to review about it

15. DateTime is no longer work, use Time always

16. There is no remote tag , use remote specifier as argument with the helper
Say for eg:

We used link_to_remote, form_remote_tag remember ?? this is going to be just
link_to, form_tag

so the code
link_to_remote("Next",{:update => "unreviewed_block"}) become
link_to("Next",{:update => "unreviewed_block", :remote => true})

and the form_remote_tag code

<%
form_remote_tag(

:update => "_results_",

:url => {:controller => '/chart/demo', :action => 'add_quick_link_refs',:id => 'link_ref'}

)

Other HTML code here

<% end_form_tag %>

becomes

<%
form_tag(

:update => "_results_",

:url => {:controller => '/chart/demo', :action => 'add_quick_link_refs',:id => 'link_ref'},

:remote => true

) do

Other HTML code here

<% end%>

Still there exists more.. will try to update if possible.

Rails 1.2.1 with Ruby 1.8.x installation tips

2012-03-30T01:21:00.003-07:00

ROR Setup (Ruby 1.8.6 , patch level 111, rails 1.2.1, CentOs 6.2 62 bit, Postgres 8.3)

Make sure rpmforge included in your yum.repo ‘s, you can make it is available through

cd /tmp/

wget http://packages.sw.be/rpmforge-release/rpmforge-release-0.5.2-2.el6.rf.x86_64.rpm

rpm --import http://apt.sw.be/RPM-GPG-KEY.dag.txt

rpm -K rpmforge-release-0.5.2-2.el6.rf.*.rpm

rpm -i rpmforge-release-0.5.2-2.el6.rf.*.rpm

Download and install ruby 1.8.6

cd /tm/

wget ftp://ftp.ruby-lang.org/pub/ruby/1.8/ruby-1.8.6-p111.tar.bz2

tar jxf ruby-1.8.6-p111.tar.bz2

cd ruby-1.8.6-p111

./configure

make

Note: If you found any error like

make

gcc -g -O2 -DRUBY_EXPORT -D_GNU_SOURCE=1 -I. -I. -c array.c

gcc -g -O2 -DRUBY_EXPORT -D_GNU_SOURCE=1 -I. -I. -c bignum.c

gcc -g -O2 -DRUBY_EXPORT -D_GNU_SOURCE=1 -I. -I. -c class.c

gcc -g -O2 -DRUBY_EXPORT -D_GNU_SOURCE=1 -I. -I. -c compar.c

gcc -g -O2 -DRUBY_EXPORT -D_GNU_SOURCE=1 -I. -I. -c dir.c

gcc -g -O2 -DRUBY_EXPORT -D_GNU_SOURCE=1 -I. -I. -c dln.c

gcc -g -O2 -DRUBY_EXPORT -D_GNU_SOURCE=1 -I. -I. -c enum.c

gcc -g -O2 -DRUBY_EXPORT -D_GNU_SOURCE=1 -I. -I. -c error.c

gcc -g -O2 -DRUBY_EXPORT -D_GNU_SOURCE=1 -I. -I. -c eval.c

gcc -g -O2 -DRUBY_EXPORT -D_GNU_SOURCE=1 -I. -I. -c file.c

gcc -g -O2 -DRUBY_EXPORT -D_GNU_SOURCE=1 -I. -I. -c gc.c

gcc -g -O2 -DRUBY_EXPORT -D_GNU_SOURCE=1 -I. -I. -c hash.c

gcc -g -O2 -DRUBY_EXPORT -D_GNU_SOURCE=1 -I. -I. -c inits.c

gcc -g -O2 -DRUBY_EXPORT -D_GNU_SOURCE=1 -I. -I. -c io.c

gcc -g -O2 -DRUBY_EXPORT -D_GNU_SOURCE=1 -I. -I. -c marshal.c

gcc -g -O2 -DRUBY_EXPORT -D_GNU_SOURCE=1 -I. -I. -c math.c

math.c:37:13: error: missing binary operator before token "("

make: *** [math.o] Error 1

Just open math.c [Don’t locate any where just use vi math.c from the same same location where you are executing the command ‘make’]

And look at the line

“#elif define(ERANGE)” it should be “#elif defined(ERANGE)”

Note: not “define”, it is “defined” => d is missing in this phrase

Modify it and again run the same command “make”

Ref: http://stackoverflow.com/questions/5193935/error-when-compiling-ruby-1-8-7-from-source-math-c37-error-missing-binary-op

http://www.iteye.com/topic/679065

Make install

Gem installation

============

cd rubygems-1.3.2

ruby setup.rb

Rails installation

============

gem install rails -v 1.2.1

it will installs the following

Successfully installed rake-0.9.2.2

Successfully installed activesupport-1.4.0

Successfully installed activerecord-1.15.1

Successfully installed actionpack-1.13.1

Successfully installed actionmailer-1.3.1

Successfully installed actionwebservice-1.2.1

Successfully installed rails-1.2.1

7 gems installed

Installing ri documentation for activesupport-1.4.0...

Installing ri documentation for activerecord-1.15.1...

Installing ri documentation for actionpack-1.13.1...

Installing ri documentation for actionmailer-1.3.1...

Installing ri documentation for actionwebservice-1.2.1...

Installing RDoc documentation for activesupport-1.4.0...

Installing RDoc documentation for activerecord-1.15.1...

Installing RDoc documentation for actionpack-1.13.1...

Installing RDoc documentation for actionmailer-1.3.1...

Installing RDoc documentation for actionwebservice-1.2.1...

Postgresql Install

=============

cd postgresql-8.1.4

./configure

If you found any error like: "readline library not found" then just try to install devel for readline.

yum install readline-devel

then do the configure again using

./configure
make

make install

.~/.bash_login or restart

env ARCHFLAGS='-arch i386' gem install \postgres -- --with-pgsql-dir=/usr/local/pgsql

Note: We need the postgres gem version 0.7.9.2008.01.28

env ARCHFLAGS='-arch i386'

wget http://rubygems.org/downloads/postgres-0.7.9.2008.01.28.gem

gem install postgres-0.7.9.2008.01.28.gem -- --with-pgsql-dir=/usr/local/pgsql

mkdir /usr/local/pgsql/data

chown postgres:postgres /usr/local/pgsql/data

ls -ld /usr/local/pgsql/data

su postgres

cd /tmp

/usr/local/pgsql/bin/postmaster -D /usr/local/pgsql/data >logfile 2>&1 &

createdb omie_dev => if needed

pg_dump omie_dev < omie_dev_backup.sql

psql omie_dev

PDF Writer

========

gem install color-tools

gem install hoe -v 2.6.1

gem install transaction-simple

wget http://rubygems.org/downloads/pdf-writer-1.1.3.gem

gem install pdf-writer-1.1.3.gem

ImageMagick

==========

yum install ImageMagick-devel

clear

[ `grep /usr/local/lib /etc/ld.so.conf | wc -l` -eq 0 ] && echo "/usr/local/lib" >> /etc/ld.so.conf

ldconfig

gem install rmagick -r --no-ri --no-rdoc

Lighttpd installation

=============

yum install lighttpd lighttpd-devel mod-lighttpd

/usr/sbin/lighttpd -v

yum install pcre-devel

yum install mod_fastcgi

yum install lighttpd-fastcgi

fcgi installation

===========

yum install fcgi

if it is not work then use the source [you need to download the fcgi with exact version]

cd fcgi-2.4.0

ls /usr/local/fcgi

mkdir /usr/local/fcgi

./configure --prefix=/usr/local/fcgi --exec-prefix=/usr/local/fcgi

make

If you found any error like:

make[2]: *** [fcgio.lo] Error 1

make[2]: Leaving directory `/somepath/fcgi-2.4.0/libfcgi'

make[1]: *** [all-recursive] Error 1

make[1]: Leaving directory `/somepath/fcgi-2.4.0'

make: *** [all] Error 2

just go include/fcgio.h and add the code “#include <stdio.h>” after

“#include <iostream>”

ref: http://stackoverflow.com/questions/4577453/fcgio-cpp50-error-eof-was-not-declared-in-this-scope

http://stackoverflow.com/questions/8833718/installing-fastcgi-dev-kit

gem install fcgi -r -- --with-fcgi-lib=/usr/local/fcgi/lib --with-fcgi-include=/usr/local/fcgi/include

Follow the openssl things at:

http://kannannadarajan.blogspot.in/2012/03/lighttpd-with-fastcgi-for-ror-121.html

install pdfkit and wkhtmltopdf

gem install pdfkit -v 0.5.0

gem install wkhtmltopdf -v 0.1.2

gem install wkhtmltopdf-binary

Make sure the following gem list is installed properly with the specific version

actionmailer (1.3.1)

actionpack (1.13.1)

actionwebservice (1.2.1)

activerecord (1.15.1)

activesupport (1.4.0)

color (1.4.1)

color-tools (1.3.0)

docsplit (0.5.0)

fcgi (0.8.8)

hoe (2.6.1)

json_pure (1.6.6)

mime-types (1.16)

parseexcel (0.5.2)

payment (1.0.1)

pdf-writer (1.1.3)

pdfkit (0.5.0)

polyglot (0.3.2)

postgres (0.7.9.2008.01.28)

rails (1.2.1)

rake (0.9.2.2)

rmagick (2.13.1)

roo (1.9.3)

ruby-ole (1.2.11.1)

rubyforge (2.0.4)

spreadsheet (0.6.5.4)

transaction-simple (1.4.0)

treetop (1.4.10)

wkhtmltopdf (0.1.2)

wkhtmltopdf-binary (0.9.5.3)

If your wkhtmltopdf is not working then use the following:

wget http://wkhtmltopdf.googlecode.com/files/wkhtmltopdf-0.9.9-static-amd64.tar.bz2

tar xvjf wkhtmltopdf-0.9.9-static-amd64.tar.bz2

mv wkhtmltopdf-amd64 /usr/local/bin/wkhtmltopdf

chmod +x /usr/local/bin/wkhtmltopdf

service iptables stop

pkill -9 ruby

pkill -9 lighttpd

/usr/sbin/lighttpd -f /etc/lighttpd/lighttpd-omie-dev-ssl.conf/usr/sbin/lighttpd -f /etc/lighttpd/lighttpd-omie-dev.conf

su postgres

cd /tmp//usr/local/pgsql/bin/pg_ctl -D /usr/local/pgsql/data -l logfile start

Lighttpd with fastCGI for ROR 1.2.1

2012-03-30T01:15:00.000-07:00

Lighttpd Installation:

As of now (18-Feb-2011) the version 1.4.28 is the current version of lighttpd, which we need yo downloaded from http://www.lighttpd.net/

tar -zxvf lighttpd-1.4.28.tar.gz

cd lighttpd-1.4.28

./configure

./configure --with-openssl

Look at lighttpd is automatically mapped at /usr/local/sbin/lighttpd , so try to execute the version check command and the output will be very similar as:

[root@kannan ~]# /usr/local/sbin/lighttpd -v

lighttpd/1.4.28 (ssl) - a light and fast webserver

Build-Date: Feb 17 2011 16:37:02

Then, note that there might be a new directory called /etc/lighttpd/ which will contains a default configuration file called "lighttpd.conf" This is the file which will contains the information about how (ROR based) web application is going to act along with script/server. This file is same as

webrick/config.rb for Webricks,
httpd.conf for Apche,
web.config for IIS etc.,

Note:

The file <RAILIS_LIB>/commands/server.rb act as a hub which will decides which server is going to serve for the specified web applications
You can refer the supported or available servers for the current rails version at <RAILIS_LIB>/commands/servers
In our case, RAILS_LIB in server is /usr/lib/ruby/gems/1.8/gems/rails-1.2.1/lib/

edit the file /etc/lighttpd/lighttpd.conf there you need to setup few of the server things

port
modules
error-handler-404
document-root
errorlog
accesslog.filename
url.rewrite
fastcgi server settings
mime type assignment

If you want any sample

Ste2 : FastCGI Library & Gem installation:

This step is entirely optional, but it will give you a performance boost, particularly when running larger Rails applications.

First you need to install the fastcgi libraries:

cd /tmp

wget http://www.fastcgi.com/dist/fcgi-2.4.0.tar.gz

tar -zxvf fcgi-2.4.0.tar.gz

cd fcgi-2.4.0

./configure --prefix=/usr/local/fcgi --exec-prefix=/usr/local/fcgi

make

make install

fcgi gem

Now you can install the fcgi gem (this is a long command which your browser may line-wrap here -- beware that the command is all on the same line):

gem install fcgi -r -- --with-fcgi-lib=/usr/local/fcgi/lib --with-fcgi-include=/usr/local/fcgi/include

Step3: Secure socked implementation

Please refer the article which is located http://kannannadarajan.blogspot.com/2012/03/why-https-and-what-is-ssl-certificate.html

There exists two certificates 1. Signed and 2. Self signed

First we will go with the self-signed certificate then we will see the signed version in our next post.

Lets see the self-signed certificate creation in UNIX:

This is done through the tool openssl, so check whether your unix environment supports the command openssl, otherwise you need to install it.

[root@kannan ~]# openssl version

-bash: openssl: command not found

[root@kannan ~]# yum install openssl

In any case you can update openssl

[root@kannan ~]# yum update openssl

[root@kannan ~]# openssl version

OpenSSL 0.9.8e-fips-rhel5 01 Jul 2008

New certificate creation

[root@kannan ~]# openssl req -new > server.cert.csr

This is a certificate request command. This will ask few question and answer it accurately about you and your site as much as possible. The main thing is PEM pass phrase. you need to remember this and the system will ask to enter when the time of certificate and private key generation by the following command.

[root@kannan ~]# openssl rsa -in privkey.pem -out server.cert.key

Self sign process

This step will differentiates the ssl is whether signed or self signed.

[root@kannan ~]# openssl x509 -in server.cert.csr -out server.cert.crt -req -signkey server.cert.key -days 365

The final feedback of the above command should be

Signature ok

subject=/C=US/ST=California/O=MNK Inc.,/OU=IT/CN=kannan/emailAddress=kannan@mysite.net

Getting Private key

Now look at in your current directory there you can se three new files called:

1. server.cert.csr

2. server.cert.crt => this is certificate

3. server.cert.key => this is private key

So, Certificate creation process is over. Now we need to tell lighttpd about this certificate.

Lighttpd: SSL configuration

create two new files files at /etc/lighttpd/kannan.net, note that kannan.net is a new directory

kannan.net.pem

kannan.net.crt

copy the contents from server.cert.crt and server.cert.key to kannan.net.pem and copy then content from server.cert.key to kannan.net.crt

A small modification in the existing lighttpd.conf file will change everything. The modification is

1. 443 is the default port for https so change it in lighttpd.conf

server.port = 443

2. Dnable the ssl engine

ssl.engine = "enable"

3. certificate and private key file location setup

ssl.pemfile = "/etc/lighttpd/kannan.net/kannan.net.pem"

ssl.ca-file = "/etc/lighttpd/kannan.net/kannan.net.crt"

Look at the attached modified lighttpd.conf which is named as lighttpd-omie_dev-ssl.conf

Thats all. stop the existing lighttpd server if anything running

pkill -9 lighttpd

Start the server

/usr/local/sbin/lighttpd -f /etc/lighttpd/lighttpd.conf

Now your browser is ready to server https version of omie_dev just type https://yoursite.com

Look at your browser raises an alert that the certificate is invalid because it is self signed. If it is signed by any CA then this will not happens.

If you want any clarification, please feel free to ask.

Thanks,

Kannan

Why HTTPS and What is SSL Certificate

2012-03-30T01:10:00.002-07:00

I want to share few things about SSL/Https/signed/self-signed Certificates. This is right time to publish about my thoughts. Lets see why https instead of http:

There are two primary differences between an HTTPS and an HTTP connection work:

HTTPS encrypts the data sent and received with SSL, while HTTP sends it all as plain text
HTTPS connects on port 443, while HTTP is on port 80, Means 443 is the default port for https, so make sure whether the the port 443 is free.

As I said above, HTTP sends the data collected over the Internet in plain text. This means that if you have a form asking for a credit card number, that credit card number can be intercepted by anyone with a packet sniffer. Since there are many free sniffer software tools, this could be anyone at all. By collecting credit card information over an HTTP (not HTTPS) connection, you are broadcasting that credit card information to the world. And the only way your customer will learn it was stolen is when it's maxed out by a thief.

There exist two kind of certificates

Signed (Provided by certificate authority like Verisign, Thawte etc., of-course you need to spend some amount
Self-signed (created using ssl tools like openssl)

Both certificates will generate a site that cannot be read by third-parties. The data sent over an https connection or SSL, will be encrypted regardless of whether the certificate is signed or self-signed, then the next question in our mind might be

Why Pay a certificate authority (CA) for Signed certification? the answer is:

A certificate authority tells your customers that this server information has been verified by a trusted source. The most commonly used Certificate Authority is Verisign. Depending upon which CA is used, the domain is verified and a certificate is issued. Verisign and other more trusted CAs will verify the existence of the business in question and the ownership of the domain to provide a bit more security that the site in question is legitimate.

The problem with using a self-signed certificate is that nearly every Web browser checks that an https connection is signed by a recognized CA. If the connection is self-signed, this will be flagged as potentially risky and error messages will pop up encouraging your customers to not trust the site.

When self signed SSL Certificate is ok

I don't believe that any site that needs a SSL certificate for customer-facing pages should use a self-signed certificate. I also don't think it's a good idea to use a self-signed certificate on any Web server that is live on the Internet. That is just asking for hackers to set up a man-in-the-middle or other hack on your server to try and trick people into providing information they shouldn't.

The only time a self-signed cerificate should be used is for testing or development behind a firewall. Such as our a kind of development environment. If it is production, we should go with signed SSL version

If you want any clarification on this please feel free to contact me.

Sample lighttpd conf file for ssl and non ssl

2012-03-30T01:05:00.004-07:00

Non SSL

# Default configuration file for the lighttpd web server
# Start using ./script/server lighttpd

server.port = 3000
server.name = "localhost"
server.modules = ( "mod_rewrite", "mod_accesslog") #, "mod_fastcgi" )
#server.error-handler-404 = "/dispatch.fcgi"
server.document-root = "/home/myapp/"

server.errorlog = "/home/myapp/log/lighttpd.error.log"
accesslog.filename = "/home/myapp/log/lighttpd.access.log"

url.rewrite = ( "^/$" => "index.html", "^([^.]+)$" => "$1.html" )
#server.document-root = "/home/myapp/"

# Change *-procs to 2 if you need to use Upload Progress or other tasks that
# *need* to execute a second request while the first is still pending.

fastcgi.server = ( ".fcgi" =>
( "localhost" =>
(
"min-procs" => 10,
"max-procs" => 10,
"socket" => "/home/myapp/tmp/sockets/fcgi.socket",
"bin-path" => "/home/myapp/public/dispatch.fcgi",
"bin-environment" => ( "RAILS_ENV" => "development" )
)
)
)

mimetype.assign = (
".css" => "text/css",
".gif" => "image/gif",
".htm" => "text/html",
".html" => "text/html",
".jpeg" => "image/jpeg",
".jpg" => "image/jpeg",
".js" => "text/javascript",
".png" => "image/png",
".swf" => "application/x-shockwave-flash",
".txt" => "text/plain"
)

# Making sure file uploads above 64k always work when using IE or Safari
# For more information, see http://trac.lighttpd.net/trac/ticket/360
$HTTP["useragent"] =~ "^(.*MSIE.*)|(.*AppleWebKit.*)$" {
server.max-keep-alive-requests = 0
}

SSL

# Default configuration file for the lighttpd web server
# Start using ./script/server lighttpd

server.port = 443

server.modules = ( "mod_rewrite", "mod_accesslog", "mod_fastcgi" )
server.error-handler-404 = "/dispatch.fcgi"
server.document-root = "/home/myapp/public/"

ssl.engine = "enable"
ssl.pemfile = "/etc/lighttpd/kannan.net/kannan.net.pem"
ssl.ca-file = "/etc/lighttpd/kannan.net/kannan.net.crt"
server.name = "your side name.com"
server.errorlog = "/home/myapp/log/lighttpd.error.log"
accesslog.filename = "/home/myapp/log/lighttpd-ssl.access.log"

url.rewrite = ( "^/$" => "index.html", "^([^.]+)$" => "$1.html" )

# Change *-procs to 2 if you need to use Upload Progress or other tasks that
# *need* to execute a second request while the first is still pending.
fastcgi.server = ( ".fcgi" =>
( "localhost" =>
(
"min-procs" => 1,
"max-procs" => 1,
"socket" => "/home/myapp/tmp/sockets/fcgi.socket",
"bin-path" => "/home/myapp/public/dispatch.fcgi",
"bin-environment" => ( "RAILS_ENV" => "development" )
)
)
)
mimetype.assign = (
".css" => "text/css",
".gif" => "image/gif",
".htm" => "text/html",
".html" => "text/html",
".jpeg" => "image/jpeg",
".jpg" => "image/jpeg",
".js" => "text/javascript",
".png" => "image/png",
".swf" => "application/x-shockwave-flash",
".txt" => "text/plain"
)

# Making sure file uploads above 64k always work when using IE or Safari
# For more information, see http://trac.lighttpd.net/trac/ticket/360
$HTTP["useragent"] =~ "^(.*MSIE.*)|(.*AppleWebKit.*)$" {
server.max-keep-alive-requests = 0
}