Unsupported Version of Google Chrome after Ubuntu update

Recently ran the 'Software Update' utility on my laptop running Ubuntu Linux 14.04 OS and ever since, I have been graced by this (un)friendly sort of error every time I use gmail app on my google chrome browser:

This version of Chrome is no longer supported. Please upgrade to asupported browser. Dismiss

This error suggests that the gmail app is not very happy with my browser version (which was v36, the latest version out there being v39). Now If I use 'Software Update' utility again, it tries to update a laundry list of all packages in my OS to update totaling upto about 600 Mb; which would be ok if I wasnt short on time. So a better solution for me would be to simply upgrade Chrome only and nothing else. And here is what I tried.

First, I tried to upgrade Google chrome via the browser settings menu but it came back and said its already at the latest! well that wasnt much help.
So I then, tried my next option of upgrading only Google chrome via the apt-get utility (as shown in the screen shot).

And that did it. Once apt-get completed its work, I restarted my browser and it on checking its version, it showed that it was upgraded to the new version (v39). And yes the error no longer shows up either! Hooray!

So, if you have been through this error yourself and wondering what to do, then you could follow the simple method of upgrading only Google chrome & get rid of that pesky error.

Installing Standalone Hadoop on Ubuntu 14.04

Today we discuss the process of Installing Standalone Hadoop on Ubuntu 14.04.

Here is the short of it.

Pre-requisites:

1- Java binaries  - Hadoop requires Java to run its process
2- Hadoop Tarball - Thats a no brainer, you need hadoop to run hadoop. :)
Note: For standalone hadoop (one node only), ssh is not mandatory.

To get Java, you can rely on the good old Open JDK (like I did myself) or you can get the 'official' java from https://www.oracle.com/java/index.html.
In my case, long time ago for some other work, I had installed openjdk (the un-official but opensource version of Java) version 7 (or java 1.7 as some like to say) according to the instructions from this page - http://openjdk.java.net/install/
We will come back to this in the configuration part later in the discussion.

As for Hadoop, you can simply download it from Apache site. I downloaded from http://www.apache.org/dyn/closer.cgi/hadoop/common/ and downloaded the (slightly older) version 1.2.1 tarball (hadoop-1.2.1.tar.gz).

The first thing to know is that you DONT need to be root to install ubuntu (as long as you have sudo).
So I was logged in as my own not-root user (mrinal)

Then I created a directory under my home directory where I moved the hadoop tarball I downloaded earlier from Apache mirror (see above) and unpacked it using the following command:

mrinal@ms-dell:~/hadoop$ mv  /home/mrinal/Downloads/hadoop-1.2.1.tar.gz   /home/mrinal/hadoop/
mrinal@ms-dell:~/hadoop$ tar xvzf hadoop-1.2.1.tar.gz

This creates a new directory in the same place and that becomes my hadoop install location.

Now, refering back to the java install process from above, in ubuntu, the java is installed by apt-get installer process under /usr/lib/jvm folder. Sometimes there may be previous older preinstalled versions of java present but you can ignore them--we are interested in the version 7.

Then I set my environment variables in my .bashrc file (for subsequent logins to be set properly). The .bashrc file resides in my home directory (in my case /home/mrinal) as follows

mrinal@ms-dell:~/hadoop$ vim ~/.bashrc 

and at the end of the file enter these lines:

Note: It is crucial that the PATH env variable includes the JAVA_HOME/bin & HADOOP_INSTALL/bin directories.

For Hadoop in distributed mode in a cluster, ssh service is required but in our case being a standalone mode, ssh is not mandatory. Nevertheless its good to know that in Ubuntu (and Linux in general) ssh is preinstalled.

The last thing before hadoop config checks is to set JAVA_HOME variable in the hadoop-env.sh file.
This file resides in the $HADOOP_INSTALL/conf directory (/home/mrinal/hadoop/hadoop-1.2.1/conf in my case). If you dont set this variable, you would get weird errors like "localhost: Error: JAVA_HOME is not set." when you try to start / stop / do anything with your hadoop.

Anyhow, once you are done with all the above steps, you should be able to do basic config checks to see if your hadoop is installed and functioning ok..

mrinal@ms-dell:~$ echo $JAVA_HOME 
/usr/lib/jvm/java-7-openjdk-amd64/
mrinal@ms-dell:~$ echo $HADOOP_INSTALL
/home/mrinal/hadoop/hadoop-1.2.1
mrinal@ms-dell:~$ java -version
java version "1.7.0_55"
OpenJDK Runtime Environment (IcedTea 2.4.7) (7u55-2.4.7-1ubuntu1)
OpenJDK 64-Bit Server VM (build 24.51-b03, mixed mode)
mrinal@ms-dell:~$ hadoop version
Hadoop 1.2.1
Subversion https://svn.apache.org/repos/asf/hadoop/common/branches/branch-1.2 -r 1503152
Compiled by mattf on Mon Jul 22 15:23:09 PDT 2013
From source with checksum 6923c86528809c4e7e6f493b6b413a9a
This command was run using /home/mrinal/hadoop/hadoop-1.2.1/hadoop-core-1.2.1.jar
mrinal@ms-dell:~$ start-dfs.sh 
starting namenode, logging to /home/mrinal/hadoop/hadoop-1.2.1/libexec/../logs/hadoop-mrinal-namenode-ms-dell.out
mrinal@localhost's password:
localhost: starting datanode, logging to /home/mrinal/hadoop/hadoop-1.2.1/libexec/../logs/hadoop-mrinal-datanode-ms-dell.out
mrinal@localhost's password:
localhost: starting secondarynamenode, logging to /home/mrinal/hadoop/hadoop-1.2.1/libexec/../logs/hadoop-mrinal-secondarynamenode-ms-dell.out
mrinal@ms-dell:~$ start-mapred.sh
starting jobtracker, logging to /home/mrinal/hadoop/hadoop-1.2.1/libexec/../logs/hadoop-mrinal-jobtracker-ms-dell.out
mrinal@localhost's password:
localhost: starting tasktracker, logging to /home/mrinal/hadoop/hadoop-1.2.1/libexec/../logs/hadoop-mrinal-tasktracker-ms-dell.out
mrinal@ms-dell:~$