I am trying to install hadoop in my ubuntu OS. I followed each and every step exactly from this link Hadoop Install Tutorial and everything was going as expected until i tried to run
$ start-dfs.sh and $ hadoop jar ./share/hadoop/mapreduce/hadoop-mapreduce-examples-2.2.0.jar pi 2 5 command. These commands doesn't work as expected.I tried R&D and somehow came to know that i was using older hadoop version Hadoop 1.0.2 despite of me getting latest 2.2.0 version.
As i could not solve this, i tried to uninstall hadoop completely, Now when i try doing it, it says
$ sudo dpkg -r hadoop
dpkg: dependency problems prevent removal of hadoop:
hadoop-native depends on hadoop (= 1.0.2-0ubuntu1~hadoop1).
dpkg: error processing hadoop (--remove):
dependency problems - not removing
Errors were encountered while processing:
hadoop
Appreciate any help !
I dont know whether its a proper way to remove hadoop or not, but i have removed it using below method.
I first manually deleted the /usr/local/hadoop folder from all the users(If any).If you are not able to remove it due to lack of permissions, then make sure about the permissions of the folder. Make the permission of the folder to "Sudo" and on "Creating and deleting files" so that every user can delete from their instances.
Then from Terminal $ rm -r hadoop does the job going to the /usr/local path.
After this, i checked $ hadoop version again in terminal ..and boom it again showed its existence. Then i did below step.
2.Goto terminal sudo apt-get purge hadoop or sudo apt-get remove hadoop...then it worked
Related
I was using lerna yarn workspaces to manage the my monorepos but due to dependencies incompatible version issues i downgraded my yarn version and after that yarn cmds start failing. Then i again upgraded my yarn version to latest 1.22.10 but after that no yarn cmd is running. running any cmd gives below error attached in SS:
Can someone help me in this, i have tried multiple times reinstalling node, nvm and yarn but nothing is working for me.
I have tried downloading yarn using brew, npm globally but nothing works
Finally, after slogging for past 2 days and searching insanely online. Able to resolve this by hit and trial.
If you have yet to install or see yarn at /.yarn/ you can install it directly following https://classic.yarnpkg.com/en/docs/install/ or
curl -o- -L https://yarnpkg.com/install.sh | bash
Then follow below steps:-
mkdir ~/.yarn/releases/
cp ~/.yarn/lib/cli.js ~/.yarn/releases/yarn-1.19.0.cjs
yarn -v
voila, it works, Never been so happy before
I am trying to install cluster hadoop on centos7. But command is not responding , I manually download the hadoop from this link on windows and then copy on VMware and installing using this command but this is not working.
What if you run:
tar xvzf /home/hadoop.2.7.0.tar.gz
Also could you please run and paste results for:
du /home/hadoop.2.7.0.tar.gz
tar tvzf /home/hadoop.2.7.0.tar.gz
md5sum /home/hadoop.2.7.0.tar.gz
You can find distribution with checksums here
Can anybody provide me with the commands to install Hadoop 2.2 on Ubuntu 14.04?
I have checked various sites but they all seem to have different procedures.
I successfully installed it using this step-by-step guide (in my case it was on 14.10, but I doubt there will be any difference)
Here is a important part:
wget http://apache.mirrors.pair.com/hadoop/common/stable2/hadoop-
2.2..tar.gz
tar –xvzf hadoop-2.2.0.tar.gz
mv hadoop-2.2.0 hadoop
sudo mv hadoop /usr/local/
sudo chown -R hduser:hadoop Hadoop
You can further configure it for your convenience, choose interface etc.
Hope this helps
I'm trying to compile/install/run Hadoop as a single node cluster on a Mac OS X 10.7.5.
I've downloaded the hadoop-2.2.0-src, and am able to compile all modules with
mvn install
The install is successful, and the tests check out too.
When trying to run hadoop (specifically, hdfs -namenode format to start off with), I see a requirement for hadoop components to exist in directories like:
/usr/libexec
/usr/lib/conf etc.
What is the install step required to get the files into this directory? Can it be done from Maven, or is there a manual install step required?
One option, and I'm not sure if it's correct, is to set my HADOOP_HOME - is this where hadoop finds its /libexecs?
Thanks guys
Pete
I set up my Hadoop clusters with Hadoop 2.0.2. Then, today I tried to test 1.0.0. So I downloaded the deb file from the Hadoop website and installed it: It did mess up everything.
Now, when I type "which -a hadoop" I get 2 results
one pointing to my old Hadoop installation folder
and the other one pointing to /usr/bin/hadoop.
So the question is: how to get rid off of Hadoop 1.0.0 completely?
Try using dpkg -r hadoop; this should remove the Hadoop package from the system, but leave the config files intact. If you want to lose the config files as well, try dpkg -P hadoop instead.
> $HADOOP_HOME
> /home/shiv/hadoop
> sudo rm -r /home/shiv/hadoop
And Hadoop is uninstalled!
I struggled through this for longer than a while and then decided to share it here:
The trick is to basically delete all the symlinks pointing back to locations where HDP components reside since that is what causes 80% of the problem. Here is a step by step tutorial for that:
http://www.yourtechchick.com/hadoop/how-to-completely-remove-and-uninstall-hdp-components-hadoop-uninstall-on-linux-system/
Hope that helps!