Today, while your organization may have a DS3 line providing WAN service, it's equally likely the provider has given you an Ethernet link throttled back to the equivalent of DS3 speed. There are several advantages to using Ethernet on the WAN for both you and your provider:
- Delivery Costs – The cost of delivery via Ethernet is less expensive than traditional delivery technologies (T1, OC3, etc.)
- Analysis Costs – The cost of Ethernet analysis solutions is less than traditional WAN analysis tools
- Familiarity – Your network administrators are already experienced with Ethernet, and won’t have to learn a new delivery technology
- Scalability – Ethernet is a more scalable technology than traditional frame-relay. This is important, as more applications continue to use greater amounts of bandwidth.
- Designed for Data – Traditional methods of WAN delivery were originally developed for transporting voice and modified to carry data. Ethernet was developed with data transportation as its primary function, making it more efficient.
You're probably wondering, "How do these changes impact the tools I use for monitoring WAN performance and traffic?" For many, it means rather than using a WAN probe appliance, you'll use an Ethernet or Gigabit probe appliance for monitoring the Ethernet connection from your provider.
Observer® Advantage
When purchasing a connection from your provider, they typically provide a full-duplex 100 Mb link throttled back to the requested speed–for example 30 Mb. Be sure your analyzer's statistics are calculated based upon the throttled speed (30 Mb) rather than the link speed (100 Mb). Observer allows you to specify the speed upon which it bases its statistics. In the main Observer console window, right-click on the probe instance and select Probe or Device Properties > Adapter Speed. In setting the speed, the link is synchronized at 100 Mb, but Observer calculates utilization statistics based upon the value you set (30 Mb).
Primary Troubleshooting Challenge
While the WAN landscape is changing, one thing remains the same: the challenge of pinpointing the source of WAN performance issues. A critical tool in your Observer arsenal for finding delay source is MultiHop Analysis.Through MultiHop Analysis you can graphically see conversations traveling across multiple network locations and pinpoint the source of delay and anomalous behavior. It also gives you proof when dealing with your provider as the source of delay. This month’s Tech Spotlight shows you how to use this.
Conclusion
Ethernet is emerging as the primary technology for WAN delivery. While Ethernet offers strong advantages in familiarity, cost, and scalability over other methods of WAN delivery, it also presents some monitoring challenges for the unaware. To learn more about how Observer Experts like MultiHop Analysis simplify WAN analysis and speed the troubleshooting process, attend one of our Thursday Observer webinars. |
- Within the Observer main console window, select Capture > Packet Capture on Multiple Instances from the overhead menu.
- From the Packet Capture on Multiple Instances Settings window, check the probe instances from which you want to simultaneously capture packets. Optionally, you can apply filters to probes by clicking the Set Filter for Select Instance button. Once the probe instances are selected, press Start.
- Both captures appear together, along with the Multiple Instance Packet Capture box. In the box, check Start MultHop Analysis. Press Stop and the captures are saved and transferred to Observer's MultiHop Analysis.
- In the MultiHop Analysis Window, click OK to move forward.
- The MultiHop Analysis screen presents all conversations witnessed by the probes. You can toggle between the Hop Summary tab for an overview of conversation statistics and the Connection Dynamics tab for an in-depth view of the connections. This detailed view allows you to find the delay source by identifying how long the packets took to travel from segment to segment.
|
