Home > Information technology essays > System architecture: NFS file system into cluster-based file system for better performance

Essay: System architecture: NFS file system into cluster-based file system for better performance

Essay details and download:

  • Subject area(s): Information technology essays
  • Reading time: 5 minutes
  • Price: Free download
  • Published: 9 December 2021*
  • Last Modified: 3 October 2024
  • File format: Text
  • Words: 1,300 (approx)
  • Number of pages: 6 (approx)

Text preview of this essay:

This page of the essay has 1,300 words.

System Architecture
1 Introduction
This is for a small company who is considered NFS file system into a cluster-based file system for better performance. The company needs to store gigabyte of private data on cloud. The main reason for the cluster base file system is to high volume of better performance. The system is established by using CentOS and VMware Workstation 10.
2 Experiment Setup and Methodology
The storage test in our experiments consists of a server and the client connected with the Ethernet LAN. Our machine is Intel Core i5 processor 2.33 GHz. The storage capacity 32GB for the work load for the server. The client is in the same machine with 2GB main memory.
• It is easy to share the file between local and remote host.
• Admin data may contain single host
• Data consumed large amount of disk space in the single host.
• It countenances the remote application
2.1 NFS file system
It is work over the IP protocol in any system TCP/IP. NFS file system is the main storage system which is called central storage system. A client requested to mount from remote host directory and in a similar way mount in a physical device.
2.2 Architecture Component
Server:
• nfsd: NFS server daemon services needs from clients.
• mountd: NFS mount daemon transfers the mount request delivered on by nfsd.
• rpcbind: RPC port mapper used to locate the nfsd daemon.
• /etc/exports: configuration file that defines which portion of the file systems are exported through NFS and how.
Client:
• mount: Standard file system mount command.
• /etc/fstab: File system table file.
• nfsiod: (optional) local asynchronous NFS I/O server.
2.3 Hardware Specification and Configuration
Name Details
Processor Intel (R) Core (TM) i5-3770 CPU @ 2.33GHz
RAM 8.00 GB (7.82 GB Usuable)
Local networking 1Gbps
Operating System Win10 64-bit
Internet Connection
Link Speed 1 Gbps internet access
Hardware firewall
Hardware and Internet Connection Specification
Information of the Hardware
2.4 Software Configuration
Software configuration is an important issue beside the hardware. We are using VMware workstation 10 with Linux based operating system which is maintenance windows 7 64 bit. A VMware workstation 10 that is Centos 7 is operating to create NFS distributed file system and check the measurement performance by using the benchmark tools.
2.5 VMware workstation 10
VMware is software that can makes a several virtual machines on a single physical machine. It has a different version of software which easily setup on different types of machine. It has support VMware cloud air virtual private on cloud. It is easy to use and safe the data by using the snapshot which back up data from the faulty hardware and software.
2.6 IOzone Benchmark Tool
It is really important to monitoring the Linux performance by using the measuring tool. In this system we use the Iozone Benchmark tool. The bench mark tool generates and measures variety of files by operations. It’s also run into the operating system. It is testing the file form I/O by the below operations.
Read, write, re-read, re-write, random read, random write, backeard read, record rewrite, stride read, fwrite, re-fwrite, fread and re-fread etc.
IOzone has 13 types of different test. It is test to the database server and makes the database performance.
By using the IOzone we get the large file system for the better performance and better balance for the operating system.
2.7 Network Topology
The network topology used for the client-server model which is always different task. The server keeps the record of the data of information for the client. Client always request the server to get the information and server always response when server knocked by the client for the data information. The server-client communicates each other form different machine.
Here we are using the Bus Network topology in a single cable for the single cable connection. It travels from the source and finds the recipient into the single cable.
Different
Hard mounts and Soft mounts
Hard mount: If the connection is lost with minor problem the entire NFS client have frozen the application, unless the NFS server comes back online.
On the other hand, after the freeze the application, it is not possible to bring it back.
Soft mount: If server fails application gets an error. The client remains responsive, but this method of mounting may lead to data corruption.
Sync Async
• Decrease lifetime of flash memory • Data did not committed immediately
• Calculated by bit to write • Write all in one time
• Data is equivalent • Data corruption is not detectable
• Server reply the client without lie anything • Server lie to the client
https://www.centos.org/forums/viewtopic.php?t=8787
TCP UDP
• Transmission Control Protocol • User Datagram Protocol
• TCP is a connection-oriented protocol • UDP is a connectionless protocol
• HTTP, HTTPs, FTP, SMTP, Telnet • DNS, DHCP, TFTP, SNMP, RIP, VOIP
• TCP is heavy-weight • UDP is lightweight
• TCP do Flow Control • UDP do not have an choice for flow control
• SYN, SYN-ACK, ACK • No handshake
http://www.diffen.com/difference/TCP_vs_UDP
NFS Implementation and Performance
Cento 7 is used inside the virtual machine to distribute the NFS file system. After install the NFS, we use the commands in both side (Server & Client) which are given below with snapshot:
Server
• [student@localhost ~] $ sudo yum install nfs-utils nfs-utils-lib
Installation of NFS
NFS Installation Completed
It shows NFS installed and we created a directory called “Lecture”
• mkdir lecture (Directory)
• ls
Mount Directory and Create File
• [root@localhost /]# sudo mount localhost:/lecture.mnt
Lecture folder has been mounted
• [root@localhost Lecture /]# Cat>>mounttest.txt
.txt file and folder has been mounttest.txt
Bandwidth Performance 1
Bandwidth Performance 2
• [student@localhost -]$ jperf-s
• [student@localhost -]$ jperf-c localhost.
Bandwidth Performance 3
• jPref 2.0.2
Sharing file in the client side
• [root@localhost /]# nano etc/exports
• /lecture10.28.10.30 (client ip address) (rw,sync)
Client Side
Installing Client side which command is given below:
• [student@localhost -]$ sudo yum install nfs-utils nfs-utils-lib
NFS Installed
Mounting on the client side
• Ifconfig
IP address are checked on client system
• Ping 10.28.10.30
Server NFS Start and server folder Mount
• root@localhost /wget http://www.iozone.org/src/current/iozone3_394.tar
• tar xvf iozone3_394.tar
• cd iozone3_394/src/current
• make
• make linux
All commands are used for benchmark toolbar.
Performances Checking
• Command line used: ./iozone -a – b
The figure shows the obdurate output values (p/s) of various tests
1st column kb: It shows the file size that is use for the testing.
2nd column reclen: It shows the record size that was used for the testing.
3rd column until the last column: It shows the several tests that are performed and its output values in per second.
Read/write, soft/hard, sync/async mounted
• root@localhost /mount localhost:/lecture mnt –o sync
• root@localhost /mount localhost:/lecture mnt –o async
• root@localhost /mount localhost:/lecture mnt –o hard
• root@localhost /mount localhost:/lecture mnt –o soft
• root@localhost /mount localhost:/lecture mnt –o rw
On the client side read/write, soft/hard, sync/async, rw has been done successfully.
DFS Parameters

  • Consistency: Consistency is to define the semantic or syntactic term. The semantic consistent the model and check the theory formula which is true or not.
  • Security: It is more important for a file system to protect the data from the bad sector and hackers.
  • Fault Tolerance: Fault tolerance is the belongings of individual system to continue operate accurately from the failure of some of its own components.
  • Transparency: Transparency is operating the way to see the actions that how it is performed.

Conclusion
We have worked in a group. Firstly we have tackled to work by Linux which is NFS file system and run into our system and solving the problems between server and client side by using coding and to enter the data into the system. After we can find out that it is really take less space and work smoothly without any errors. It’s really reliable for the company.
2015-12-13-1450032370

About this essay:

If you use part of this page in your own work, you need to provide a citation, as follows:

Essay Sauce, System architecture: NFS file system into cluster-based file system for better performance. Available from:<https://www.essaysauce.com/information-technology-essays/system-architecture-nfs-file-system-into-cluster-based-file-system-for-better-performance/> [Accessed 20-12-24].

These Information technology essays have been submitted to us by students in order to help you with your studies.

* This essay may have been previously published on EssaySauce.com and/or Essay.uk.com at an earlier date than indicated.