Application Traffic Management
1.1 Overview
With the exponential growth of bandwidth-intensive applications, not only the organizations but also the network service providers are facing various challenges in optimizing the application performance. Major issues demanding attention are: need of differentiated upstream and downstream bandwidth, network capacity, optimization needs for VoIP and IPTV, and the high customer-expectation for the network availability.
A significant portion of the application traffic comprises multimedia file sharing which are often hundreds of MBs in size, such consistent traffic load creates network congestion, making it difficult for the operators to get effective network planning.
An efficient management of the application traffic can empower the operators to manage multiple applications and offer the performance what the subscribers desire.
Table of Contents
- Application Traffic Management
- Application Traffic Types
- Tips for Monitoring Traffic on your network
1.2 What is Application Traffic Management?
Also known as the network traffic management, it refers to the process of monitoring and controlling the activities related to the network and its availability. It aims at strengthening the capacities and transforms the network into a valuable and managed resource by enhancing the efficiency, functionality, and security.
1.3 Benefits to the businesses
As discussed, a better managed network has several leading edges for the organizations. It improves the quality of services rendered, better access to digital content, easy centralization of control for multiple applications and sites, better security, and greater productivity.
2. Application Traffic Types
2.1 Bursty Traffic
Bursty Traffic refers to the inconsistent type of traffic on the network, which occurs in the form of burst at a moment and then calms down. Examples of such sudden traffic peaks include FTP downloads, graphics, videos and other large contents.
The problem with bursty traffic is that it immediately consumes a high bandwidth and suddenly exhausts the application availability. Such traffic can be efficiently managed by limiting the access to a bandwidth.
2.2 Interactive Traffic
In this type of traffic, there are sessions of short requests and response pairs. Applications are involved in real-time interaction with an end user. Web browsing, e-commerce purchases, SSL transactions, Tele-net sessions are examples of Interactive Traffics.
The problems associated with interactive traffic are the susceptibility of bandwidth to the competitors and also results in poor response time of application. Prioritizing the requirements over less essential traffic is the solution for managing this traffic.
2.3 Latency Sensitive Traffic
This is the type of traffic in which the operator has the responsibility of delivering the service on time. It is a steady stream of traffic and the examples are VoIP, Video Conferencing, live gaming, video streaming, etc. There is a sudden rise in traffic like the bursty traffic but the required data packets are generated at a regular time interval.
The problem with this is the susceptibility of bandwidth for the competitors and may result in poor application performance. Setting the bandwidth range on the basis of priorities is key here.
2.4 Non-Real-Time Traffic
This type of traffic is generated by the application requests in which the real-time delivery is not so important. Batch processing applications, Emails are examples of such traffic.
The only problem with this traffic is the consumption of bandwidth during business hours. So, scheduling the bandwidth during non-business hours is key to effective management.
3. Tips for Monitoring Traffic on your network
Managing the application traffic is important to ensure the peak performance and better security. The information acquired through various sources can be helpful in making the application run smoothly. The following tips will be greatly helpful in meeting the user-expectation.
3.1 Selecting the Right Data Sources
There are two main data-sources:
1. Flow Data – acquired from layer 3 devices, like the routers
The flow data helps in gaining information about traffic volumes and journey of network packet from origin to destination. This helps identify the unauthorized WAN traffic and utilize the available resources for better performance.
2. Packet Data – sourced from SPAN, mirror ports, etc.
This data makes the admin understand the interaction of the user with the application and track their usage on WAN. Suspicious malwares or other security concerns can also be assessed with these data sets.
3.2 Choosing the right points on network
Many times the businesses make the mistake of including too many data points while deploying the network traffic assessment tool. In actuality, there is not much need to assess each and every data available on the network. One must start with the internet gateway monitoring, which is an excellent source of security and operational data. A SPAN or the mirror port can also be configured at the network core, allowing the capture of data passing through it.
3.3 Not settling only with Real-Time data
Real-time data is a great resource for an effective monitoring of the traffic, but in reality, it is not the ‘panacea’ information. For the most optimized performance, historical data also holds a great relevance as it helps in identifying the changing trends, analysis of past events, and comparing the network activities with the past performances.
3.4 Associating data with username
Traditional tools of network monitoring generally report the activities using IP or MAC address. But, if one tries to assess a problematic device in a DHCP environment, the data with IP or Mac creates problems in getting the correct information.
Usernames can bring together the devices and activities and this association will let one assess what a user is doing on the network.
3.5 Flow and packet payloads
Though the majority of the network configurations have an intrusion detection system but most of them lack the monitoring of internal traffic. A corrupt mobile or rogue IoT device in the internal system can bring the network security to question. Also, misconfiguration can allow the firewall permit malicious traffic as well. Therefore, internal monitoring is essential.
