Solving Performance Problems Series: Easy Steps for Setting up PathView Cloud on your Network by Team AppNeta October 29, 2012
How Do I?
This is the first of a series of posts detailing how to solve common network and application performance issues and to get the most out of PathView Cloud. To solve any problem you need to have the network instrumented properly, so today we will cover:
How do I set up PathView Cloud on my network?
1. Understand your topology
Before you begin any setup, you should have an understanding of the size, scope and topology of your network. This includes identifying:
- Locations where applications are served – both internal and hosted services
- Locations where applications are consumed
- Applications being delivered
- Connectivity between these locations, including locations with redundant network connections
- Connectivity options for LAN and Wifi at remote locations
2. Place appliances in appropriate locations
In order to have proper performance visibility, you need to have an infrastructure in place. A very common discussion we have with customers getting started with PathView Cloud is the value of having just a central monitoring appliance vs. having appliances centrally AND at remote locations.
- Single-Ended vs. Dual-Ended Path Monitoring
PathView is different from pretty much every other performance monitoring solution available today because it can measure the end-to-end performance without having devices, agents, probes or any other synonym you can think of for some software running on both ends of the service delivery path. With just a single PathView appliance, you can measure performance to any IP Addressable device you can ping.
This single-ended capability is critical and is the best method on the market to understand performance to services where you don’t own the remote infrastructure, like Saas services.
But if you do have the option to have technology installed at the remote location, you get a substantially better view of performance with what we call dual-ended monitoring.
As you can see from the screenshots below, both single-ended and dual-ended monitoring are showing degraded network performance, but the dual-ended path shows that the loss is only occurring in one direction, from Boston to San Francisco, and the return leg is clean.
You will also notice that the single-ended path has an area colored red, which indicates an SLA violation. While both single-ended and dual-ended paths have SLA violation alerts, currently only single-ended paths have diagnostic test capability.
While we have designs on resolving these trade offs today, the best practice is to have both a single-ended and dual-ended path created where possible.
Now that you know the overall structure of you network and understand the key locations, it’s time to setup PathView to monitor performance. We have made significant strides forward in this area in recent months, taking the process of instrumenting a large network from laborious to easy (and dare I say it – kind of cool!).
- To get started, click the button from the overview dashboard or path list page.
- Next choose the appliances you have installed in your organization. Clicking the check box in the top left corner of the select appliance table will automatically choose all appliances and then click the button.
- Then, choose if you are monitoring out to other PathView appliances or other IP addressable end points. Choose “Set Appliances as End Points” and an interactive display of your appliances will be shown.
This display shows your PathView appliances and the Paths to be created between them to instrument your network. With the controls on the left, you can choose your network topology of Mesh or Hub and Spoke and the diagram update showing you the layout. In Hub and Spoke you can choose the core network, and in both topologies you can choose single-ended, dual-ended paths or both.
Mesh Network Setup
Hub and Spoke Network Setup
4. Monitor your Selected Targets
By selecting “Choose Targets” you can enter the host names or IP Address of hosts and services, and a path will be created from each of the selected appliances to these targets.
The final step in both the automesh and the remote target monitoring is that you can view the final detail, edit any settings or cancel the creation of any individual paths. Once you have your network instrumented properly, you can start to answer the more pressing questions about network performance, such as:
Upcoming Topics in this Blog Series
- Where is all of my bandwidth going?
- Why is my WiFi slow?
- How do I troubleshooting poor web application performance?
Please email me at email@example.com with any questions on about solving performance problems or other topics you would like covered in this series. Thanks for reading and good luck getting the most from your network.