MySQL - Create a new user

Connect to MySQL server - 

mysql -u root -p -h hostname

Execute the below command from mysql client

CREATE USER 'user_name'@'%' IDENTIFIED BY 'pwd';

GRANT SELECT, INSERT, UPDATE, DELETE, CREATE, ALTER, INDEX  ON db_name.* TO 'user_name'@'%';

We can add more privileges  like DROP if the user/application needs to drop any tables.

MemSQL - HipHop for SQL

MemSQL is a new database system developed by ex-facebookers to speed up relational databases. As its name implies, it achieves fast performance by keeping data in memory by removing the most common bottleneck "disk" which most applications hit today. Also it supports MySQL client for querying so it doesn't need you to learn anything new and you can easily migrate from MySQL to MemSQL without any change in your application code.

MemSQL takes a lesson learned from HipHop (facebook's tool for converting PHP code into faster C++) and converts SQL to C++. Unlike traditional relational database systems which interpret SQL queries, MemSQL transforms SQL to C++ and compiles the generate code into machine code. This is only one time operation when a query hits first time, and all the future queries which follow the same skeleton (with all the numeric and string parameters stripped off) will bypass code generation and the compiler entirely.

For example, when you execute a query
  SELECT * FROM users WHERE id = 10 and name = "prasad",
it is first scanned through a linear scan parser which strips of all the numeric and string parameters and creates a query skeleton. The resulting query skeleton is just a string and looks like
 SELECT * FROM users WHERE id =@ and name = ^.
This query skeleton is given to code generator which converts it into C++ code that executes the query. The resulting plan is loaded into the database as a shared object and registered in the plan cache.
And all the future queries which match the same skeleton will be directly executed without going through code generation and compilation process.

Having said that MemSQL uses RAM as the primary storage of data, it backs up data to disk with snapshots and transaction logs to not to lose any data on system restart or failure. So your data is durable. These features can be tuned all the way from synchronous durability (every write transaction is recorded on disk before the query completes) to purely in-memory durability (maximum sustained throughput on writes). 

So why wait? I have installed in MemSQL in my 64-bit linux machine today and connected it to using phpmyadmin within 10 minutes. 


Click here to download the developer edition and install it by yourself. 


~ Cheers

Ubuntu Security updates

Executing the below commands installs all updates ( includes security updates also)
sudo apt-get update
sudo apt-get upgrade

There is a package unattended-upgrades, which provides functionality to install security updates automatically.

You could use this, but not configure the automatic part and call it manually instead, so the following should do it:
sudo unattended-upgrade  

(assuming the package is installed by default, which probably is)

See also /usr/share/doc/unattended-upgrades/README.

AWS - Delete AutoScaling

1. Update your active auto scaling group with miz-size 0 and max size 0 with the below command. You should have AWS auto scaling command line tools installed.  You can follow my blog post  to setup command line tools.
$ as-update-auto-scaling-group prod-scaling-group --min-size 0 --max-size 0

2. Wait for the instances to shutdown.

3. And now delete your auto scaling group
$as-delete-auto-scaling-group pd-prod-scaling-group

Upgrade NGINX server on Ubuntu

sudo service nginx stop

sudo add-apt-repository ppa:nginx/stable

sudo apt-get update  

sudo apt-get install nginx

Configuration file `/etc/nginx/sites-available/default'   ==> Modified (by you or by a script) since installation.   ==> Package distributor has shipped an updated version.     What would you like to do about it ?  Your options are:      Y or I  : install the package maintainer's version      N or O  : keep your currently-installed version        D     : show the differences between the versions        Z     : background this process to examine the situation   The default action is to keep your current version.  *** default (Y/I/N/O/D/Z) [default=N] ?

Choose Option = N

 reboot

Amazon RDS - MySQL - Turn on Slow Log and General Log

Following commands create a new db parameter group for your RDS server and set 
slow_query_log and general_log parameters


You will need to setup RDS tools configured with your keys to execute the below commands


$ rds-create-db-parameter-group slow-log-group -f mysql5.1 -d "This is created to enable slow query logging and a few other db parameters"

$ rds-modify-db-instance pd-rds-server --db-parameter-group-name slow-log-group --apply-immediately

$ rds-modify-db-parameter-group slow-log-group --parameters "name=slow_query_log, value=ON, method=immediate" --parameters "name=long_query_time, value=1, method=immediate" --parameters "name=min_examined_row_limit, value=100, method=immediate" "name=general_log, value=ON, method=immediate"

$ rds-describe-db-parameters slow-log-group

$ rds-reboot-db-instance pd-rds-server 

Setting up git client for AWS EBS environment in windows machine

Prerequisites

1. Git should be installed

2. Powershell 2.0 should be installed

3. AWSDevTools-OnetimeSetup.bat must be run.

Setup

1. Create a folder

2. Git Init in the  folder

3. Copy AWSDevTools-RepositorySetup.bat to the current folder

4. Run AWSDevTools-RepositorySetup.bat

5. Delete AWSDevTools 

3. Open Git Bash and cd to the current folder

4. Execute aws.config 

 - Enter AWS credentials

 - Enter AWS Beanstalk application name

 - Enter AWS Beanstalk environment name

5. git add . 

6. git commit -m " Notes here "

7. git aws.push

Openfire chat server installation

Execute the below command to setup Openfire on ubuntu system:

   1  sudo add-apt-repository ppa:sun-java-community-team/sun-java6 
   2 sudo apt-get update
   3  sudo apt-get upgrade
   4  sudo apt-get install sun-java6-jre
   5  sudo wget http://www.igniterealtime.org/downloadServlet?filename=openfire/openfire_3.7.1_all.deb
   6  sudo mv downloadServlet\?filename\=openfire%2Fopenfire_3.7.1_all.deb openfire_3.7.1_all.deb
   7  sudo dpkg -i openfire_3.7.1_all.deb

On Redhat, CentOS, use the below commands

  1 sudo wget http://www.igniterealtime.org/downloadServlet?filename=openfire/openfire-3.7.1-1.i386.rpm

  2 rpm -ivh openfire-3.7.1-1.i386.rpm

- Open http://yourserver.com:9090 from your browser and complete setup process such as db integration, domain etc.

Education System

Our education system has to bring lot of changes in rural areas. There are many students out there who could not afford to get quality education and do not get support from their parents to go to schools.

1. Free education: There should be free and "quality" education at least till the school level. It's definitely welcomed if we can make entire education to the students free.

2. No Reservation: Instead of applying reservation for getting seats in any professional course (especially medicine type of courses), we can provide free education for students whose parents' income is low (irrespective of caste) and based on merit.

3. Govt schools and colleges: In most of the govt schools and colleges, in rural areas, though we have skilled and good qualified teachers results are not very good. Most of the parents tend to join their children in private institutions where you may not find qualified teachers (and quality education as well) but their children get good percentages or whatever. Can we change the "typical attitude" of  govt institutions' staff who do not show interest in students career and feel that they are helpless in changing the current system? Probably we can have a system which 

• Monitors the students results in all the govt educational institutions and take actions to find the causes.
• Gives sort of bonuses or awards to schools or staff who work hard in getting good results of the students. 
• Have a compliant box where student or staff can give complaints or suggestions for any improvements and those complaints should very regularly be checked. 

Setting up ejabberd clustering

Machine 1: 

1. Install ejabberd as a cluster node on machine 1 (Node name: ejabberd@PrasadK)

2. Configure it to use mysql database based on the installation guide.

3. Started the server using windows services. (You can alternatively start it from command line using $ejabberd-installation/bin/ejabberdctl start)

Machine 2: 

1. Install ejabberd as a cluster node on machine 2 (Nodename: ejabberd@riktam-acb56375)

2. Configure it to use mysql database (the same server which the first machine's ejabberd server is connected to) 

3. Copy the .erlang.cookie file (which resides in C:/Windows folder if it's started as a service or $USER_HOME folder if it's started via command line) from the machine 1 to the machine 2's home folder. (I copied it in two places - in bin folder where erl command is executed and the user's home folder. Need to confirm which of them is used)

4. Execute the command from $ejabberd-installation/bin/ directory

D:\ejabberd\bin>erl -sname ejabberd@riktam-acb56375 -mnesia extra_db_nodes "['ejabberd@PrasadK']" -s mnesia

5. Executing the above command should enter into erlang shell which looks like:

Eshell V5.6.4  (abort with ^G)

(ejabberd@riktam-acb56375)1> 

         a) execute mnesia:info(). to check if the two nodes are shown in running_nodes. If not check your ejabberd.log file:

(ejabberd@riktam-acb56375)1>mnesia:info().

---> Processes holding locks <---

---> Processes waiting for locks <---

---> Participant transactions <---

---> Coordinator transactions <---

---> Uncertain transactions <---

---> Active tables <---

schema         : with 29       records occupying 3580     words of mem

===> System info in version "4.4.5", debug level = none <===

opt_disc. Directory "d:/ejabberd/bin/Mnesia.ejabberd@riktam-acb56375" is NOT use

d.

use fallback at restart = false

running db nodes   = [ejabberd@PrasadK,'ejabberd@riktam-acb56375']

stopped db nodes   = []

master node tables = []

remote             = [acl,caps_features,captcha,config,http_bind,iq_response,

                      irc_custom,local_config,mod_register_ip,motd,motd_users,

                      muc_online_room,muc_registered,muc_room,passwd,

                      pubsub_index,pubsub_item,pubsub_last_item,pubsub_state,

                      pubsub_subscription,reg_users_counter,route,s2s,session,

                      session_counter,sql_pool,sr_group,sr_user]

ram_copies         = [schema]

disc_copies        = []

disc_only_copies   = []

[] = [sql_pool,local_config,caps_features,mod_register_ip]

[{ejabberd@PrasadK,disc_copies}] = [config,pubsub_subscription,passwd,

                                    irc_custom,sr_user,motd,acl,pubsub_index,

                                    sr_group,motd_users,muc_room,pubsub_state,

                                    muc_registered]

[{ejabberd@PrasadK,disc_copies},{'ejabberd@riktam-acb56375',ram_copies}] = [sche

ma]

[{ejabberd@PrasadK,disc_only_copies}] = [pubsub_item]

[{ejabberd@PrasadK,ram_copies}] = [http_bind,reg_users_counter,

                                   pubsub_last_item,route,s2s,captcha,

                                   session_counter,session,iq_response,

                                   muc_online_room]

3 transactions committed, 0 aborted, 0 restarted, 0 logged to disc

0 held locks, 0 in queue; 0 local transactions, 0 remote

0 transactions waits for other nodes: []

ok

b) Now run the following in the same `erl' session:

(ejabberd@riktam-acb56375)2> mnesia:change_table_copy_type(schema, node(), disc_copies).

c) Now you can add replicas of various tables to this node with `mnesia:add_table_copy'

or `mnesia:change_table_copy_type' as above (just replace `schema' with another table

name and `disc_copies' can be replaced with `ram_copies' or `disc_only_copies').

d)  Run `init:stop().' or just `q().' to exit from the Erlang shell. This probably can take 

some time if Mnesia has not yet transfered and processed all data it needed from first.

6) Replace the newly created Db folder with the actual DB folder in (D:\ejabberd\database) 

7) Change http_bind storage type to RAM and disc copy from Remote copy (Need to check whether it can be done in step 5.c. itself)

References:

Installation guide

http://www.ejabberd.im/node/517

http://www.blinkenlights.ch/ccms/linux/ejabberd.html