BN208 Networked Applications

Question:

This assignment is intended to provide deeper analysis of various distributed network conditions.

This assessment should be completed by students.

a)A variety of technologies and techniques are used to manage, deploy and optimize networked applications.

b. Describe the most important networksed application architectures or technologies.

c. Analyse performance issues and deployment issues related to networked software. Use industry-relevant tools, techniques and methods for managing networked software.

Answer:

Introduction

The access method for communication is sometimes referred as protocol. We define protocols as the standards for defining data exchange over networks such as the Internet, Local Area Network (LAN), and Local Area Networks (LAN).

Every protocol has its own purpose and method, as well as the specific format that governs the data exchange between the parties.

TCP and HTTP are the most commonly used protocols.

These protocols allow data to be communicated over the internet or World Wide Web.

The communication protocols are the means by which communicating parties can exchange information in the form of data packets. They can also be used to link any physical quantity.

These rules can be used to communicate in any situation. They include syntax, synchronization, semantics, and error recovery functions.

You can implement the protocols over hardware or on software. Most often, it is a combination.

The defined formats are used by communicating parties to exchange messages. Each message is unique and has a defined meaning.

The way it is implemented does not affect the behavior.

The protocol must be agreed to by all the parties.

Analyze the protocol in terms of QoS, overall network performance and overall performance

It is an open-source protocol used to access centrally stored information over the network.

Although it is based on the X.500 protocol for directory sharing, LDAP uses less resources than X.500. Hence it is sometimes called “X.500 Lite”.

LDAP organizes all files and information as in X.500.

These stored directories are available as Network Information Service and have greater storage capabilities than ever before.

This allows you to access your account from anywhere on the network with the LDAP enabled.

In many cases, LDAP can also be used to create a virtual phone directory. This allows the user to see the contact information for other clients via the network.

LDAP, however, is flexible and can query all servers that have enabled LDAP. It provides an ad-hoc database of all users globally.

LDAP is still used by many organizations around the world, including government and private universities.

LDAP, a standard directory service, is available to all users who use the internet for worldwide queries and intranets for inter-organisation based queries. This allows them to retrieve information from multiple vendor based directories.

The networking protocol allows you to query and modify multiple directory-based services that are running on the TCP/IP.

The latest version LDAP, version 3, is now available. This version allows the user to control hubs, routers and switches.

This interface can be used for writing information to the directories.

This protocol can be used by Equipment vendors to create and update their directory information.

The current version RFC 3377 of the LDAPv3 ETF is available.

Performance of the LDAP network

The LDAP’s network performance varies according to the different operations performed across directories.

While performance with lower load on the server is acceptable, it can cause performance dips when there are higher server loads.

The LDAP performance has been poor for query types related to updates and add. This is because of the lock and write operations that the LDAP performs and the resource being updated to revert any new directory.

However, this is not the case when the server is being queried to execute a search-based query. This is because the server doesn’t need any lock over data and a simple query is being run from the central storage.

NFS Protocol Family

This protocol is based on the distributed file protocol that Sun Microsystems developed in 1984. This protocol allows users to access the network storage resource just as they would local storage.

It is built on the Open Network Computing Remote Procedure Call System (ONC RPC).

NFS, the open standard described in the RFC (Request For Comments), allows anyone to create applications based upon network directory services. They can modify the NFS if they need it.

The Policy Enforcement Point (or PEP) is the central point that allows the network device or policy to enforce the rules on the network.

It could reside on any server, client, or other working node located between the paths of client and client.

LPM (Local Policy module) is a module that enforces the policy on network devices. It also controls devices that can be policy aware.

RSVP modules ask for the LPM to get the RSVP message receipts and make a decision based upon the instructions.

The PDP (policy-driven point) acts as a logical interpreter. This interprets policies on the network, and forms RSVP requests and responses that can be sent over the network.

The PDP determines who gets the attention regarding QoS and from which node to which end node, using what path.

As mentioned, the PDP takes decisions based upon administrative policies.

PEP/PDP are typically installed on the same computer, but can be moved to other machines.

You will have to address the issues of consistency and scaleability if the PEP/PDP are enabled on the same computer.

The policy server must be separate to support multiple PEPs-based machines.

PDP’s are useful for network monitoring. With all data passing through the central hub, the central headquarters manages the services and serves the QoS.

This is usually a sign that the file server doesn’t have any client information, and there are no NFS-based connections.

NFS, for example, would store the information needed to open a file. However, NFS stores the file handle as the file handle. These file handles, which are unique in nature, are immutable identifiers to each file.

Different handlers can create uniqueness by using the i–node number (or disk blocks).

To proceed with the file read operation, the client must first set up the read procedure. Otherwise, identical files would produce the exact same results.

RPC performance is only one-sided.

NFS performance in is poor over WAN but excellent over Faster LANs.

Appropriateness in Finding

The results show that protocols function as expected and have all the required architecture and procedures for messaging over the network.

All routes and procedures are as per the RFC.

Conclusion

The LDAP and NFS can be interpreted in two ways:

The LDAP protocol, which is used for directory listing, is extremely efficient. It communicates directory information over a network to which the LDAP based computer does not require any prior information.

The NFS protocol, which is currently being analyzed and observed over RPC protocol over TCP exchange required messages over the network.

It was also discovered using literature surveys that the NFS is very slow over the WAN but quick on Local Ethernets.

Refer to

[1]J. Cortes. “A brief intro to LDAP’s future”, IEEE Distributed Systems Online. vol.

[2] RFC 7612 – Lightweight Directory Access Protocol LDAP : Schema to Printer Services”, Tools.ietf.org. 2017, 2017.

[3]V. Koutsonikola & A. Vakali. “LDAP framework, practices and trend”, IEEE Internet Computing, vol.

An, Foundations and intelligent systems, 1st Ed.

[5] S. Guo and W. Yang and G. Wang “NFS Protocol Performance Analysis and Improvement for Mobile Transparent Computing”, IEEE 10th International Conference on High Performance Computing and Communications & IEEE International Conference on Embedded and Ubiquitous Computing 2013.

[6]B. Li, P. Zhang, Z. Huo, and D. Meng, “Early experiences with Write-Write Design for NFS over RDMA”, IEEE International Conference on Networking, Architecture, and Storage, 2009.

[7]Weikuan Y and J. Vetter “Initial characterizations parallel NFS implementations”, IEEE International Symposium on Parallel & Distributed Processing Workshops and Phd Forum, 2010.