[ProTechSkills]

For more information on SSH service, you can refer to the official website @
http://www.openssh.com/

For more details on how to check SSHD service status, you can read a blog @
http://www.ewhathow.com/2013/09/how-to-check-if-ssh-is-running-on-linux/

[ProTechSkills]

Hi Shivam,

This is regarding the WinScp connection and installation of openssh-server that you mentioned in your comment. The WinScp uses the SSH service and connects to 22 port. If your connection is successful without installation of openssh-server, then you can assume that the package is already installed.
To verify the same, you can execute netstat command like:

netstat -nl | grep :22
OR 
sudo netstat -nlp | grep :22

The output of the above command will look like:
tcp        0      0 0.0.0.0:22              0.0.0.0:*               LISTEN      -               
tcp6       0      0 :::22                   :::*                    LISTEN      -

You can also verify the list of SSH commands. Write ssh on your terminal and press double tab. This will list the commands starting with ssh like:
$ ssh
ssh            ssh-agent      ssh-copy-id    ssh-import-id  ssh-keyscan    
ssh-add        ssh-argv0      sshd           ssh-keygen     ssh-vulnkey

Hope the information useful.

Thanks,
Tutorial Village

• This reply was modified 9 years, 2 months ago by ProTechSkills.

[ProTechSkills]

I Guess these folders are there as part of Replication process,

>”rbw” sub directory to DataNode to persist RBW (ReplicaBeingWritten), RWR(ReplicaWaitingToBeRecovered), and RUR(ReplicaUnderRecovery) replicas.

The directory “current” has two subdirs:

“finalized” has all finalized replicas and
“rbw” has all replicas being written to.

> when datanode restarts, it reads replicas under “finalized” and add them to the replicasMap as FinalizedReplica.

> It also reads replicas under “rbw” and adds them to the replicasMap as ReplicaWaitingToBeRecovered with its length set to be the number of bytes that match its crc.

More we should ask the champion : SHREY…. Where are you man????

[ProTechSkills]

NFS refer to Network File System. It provides a shared directory to a number of machines that can access the directory as good as a directory on local file system. On the other end, HDFS is a distributed file system, in which different machines grouped for storing the data in a distributed way.

Following are some of the key differences:

1. In NFS, All data belonging to an entity(file or set of files) stored on a single machine and thus require a dedicated hardware. In HDFS, the same data is stored in a distributed way on set of commodity hardwares.
2. HDFS allows parallel processing of data through MapReduce. Since data on NFS is stored on a single location, it has beed read sequentially.
3. HDFS replicates the data and thus provide fault tolerance. Since data in NFS is stored on a single machine, it is difficult to access/restore if the machine goes out of network.

Hope this would help.

[Author]

Posts