Let’s say you are a CIO of a company with proven WAN infrastructure and are receiving complaints from the users….. the fact that applications are sluggish. Do you want to contact the network dealer (e. g. Cisco, Juniper), invest in WAN accelerators (e. g. Packeteer, Riverbed), and hire a consulting agency to investigate? What will you do to fix the situation?
To start…. take a step back and take a deep breath. You’re often jumping the gun suggesting the WAN is responsible. More information should be applied.
Table of Contents
The moral of the report…. get data first.
Determine the performance of distinct actions when the client runs in the same building as the application server.
Repeat this for a client in various locations and compare the outcome.
This may point to specific WAN links to check. In other words, understand the bandwidths and response times all over different segments. [Are some locations longer than others, does distance often play a part, and is there one distinct congestion point? ]
This may also point out the need to analyze your client traffic flow. Does the application involve 100+ round trips between your client and server (faster than the WAN) for the rest of the simplest of actions? This type of thing can only be solved using changing the application. (Don’t health care who the providers usually are. You can’t do much about the speed of light around the globe for that many moments. )
So part of the response is…… asking any distinct vendor to “fix” the item before you know what “it” is definitely will waste your time.
Often the short answer is….. find data, analyze it, and fix it.
Performance problems are usually attributable to the network, the servers, the database(s), and the applications. Hence, it is essential to look back at the complete infrastructure, including the purposes and all their components, before assuming the WAN is a culprit.
Ask yourself……
– Will be the current “end-user experience” for all the businessmen applications in terms of performance in addition to availability?
– What is the recent response time contribution connected with client, network and web server tiers?
– What are the current resource utilization levels for the critical servers that help support the business?
– What is the recent utilization of network resources (i. e., WAN links), in addition to which applications are using the most bandwidth essentially?
– Which hosting space, workstations and business destinations represent the “top talkers” on the network?
If you feel that you already have these questions, your primary suspect is WAN performance….. you need to consult whether you have sufficient bandwidth for the application traffic spanning the WAN and regardless of if the problem applications are suitable for WAN deployment in the first place. (A chatty 2-tier database app will never scale well across the WAN no matter how much h/w you throw at it).
A multilevel monitoring tool that can help produce a comprehensive network assessment within a period be it peak traffic, a 24hr workday or busy business time, can help you understand the worst offenders. Once you know who they are, you can do certain things…..
– look at the network ability for the application(s)
– consider the applications themselves to determine whether or not they are optimized for your surroundings.
If the issue is just bandwidth, the question becomes simply how much more do I need? A community profiling tool with predictive capabilities can help you assess the influence of network changes in application performance regardless of whether more bandwidth or lowered latency will solve the situation.
If latency is a concern, you may look at the problematic applications in question to see whether they are suitable for WAN deployment. The same profiling capability mentioned above can help you determine the effects of fewer round-trips between client and server(s), allowing you to determine the particular cost/benefits of changing the application or the infrastructure.
Don’t merely throw accelerators at that, however, without first realizing the problem. They may be a waste of money for specific software and not solve the problem.
It is a typical issue for most agencies at some point, and I would stick to the steps below to resolve that.
First, define and assess the problem. Poor WAN program performance can result from a lack of bandwidth, failing community gear, telecom vendor concerns, poor application design, and unforeseen network demand. The symptoms of the problems need to be documented. Could it happen at a particular time frame of the day, or is it continual? Does it occur when several applications are running? Which users will be the effect? What exactly nodes on the WAN usually are affected? Is it a localised issue, or does it often affect several locations? Include improvements made to the WAN not long ago (new hardware, new purposes, new telecom vendors, etc.).
To be able to define the issue, you should start collecting good facts to help in the process. Places to get started collecting information would include things like:
– Help desk. A good source for defining symptoms of the challenge.
– Network Gear. Routers, Switches, and CSU/DSU logs are usually reviewed for problem detection.
– Telecom vendors. For shared networks (frame pass-on, atm, etc.), they will be competent to provide statistics on broken open rates and usage.
Instructions User Interviews. Some end users don’t log all their complications.
Through these sources, you will be able to characterize the problem. The nature of the problem may dictate the solution. Some difficulties and solutions would contain…
– Poorly Designed WAN Application. Possible solutions are usually; re-work the application, moving hosts in network topology, boosting WAN bandwidth or making use of terminal server software (i. e. Citrix). The group aid and no-brainer method uses Citrix and takes WAN out of the equation.
: Poor Telecom Vendor help. Possible solutions are transforming our Telecom vendors and having them re-engineer the links. Like moving from Frame Get across to Point-to-Point topology. Yet be careful. Most of the time, the last distance is usually the same physical moderate, which may be the problem. Meaning altering the topology would not help.
Rapid Limited Bandwidth. The probable solution is increasing bandwidth. Often natural organizational growth along with uses of the WAN is the cause of the poor WAN performance and buying more bandwidth. Or maybe, once again, you can limit this kind of growth by falling back into using Citrix or maybe shifting servers.
– Failing networking gear. The possible answer would primarily be sometimes replacing components or the total piece of equipment. This is the one condition you might want to consult with an outside pro, depending on the level of skill you could have internally.
– Odd conditions. There are always strange situations. For example, I have viewed WANs slow down when people commence to email around MPEGs. It is more of a policy issue.
To put it briefly, sometimes you may have to bring some sort of consultant in. But in many situations, you may determine the cause and answer to the problem internally by using common sense.
Read also: Employ Anonymous Proxy To Keep Peeping Internet Service Providers From Viewing You
