WizShark Beta: Visual Wi-Fi Troubleshooting in the Cloud
WizShark is changing Wi-Fi troubleshooting for the better. This tool is now open for private beta; contact us to join.
At Interop in Las Vegas, a Wireless Field Day delegate and all around wireless guru George Stefanick presented “The Wireless Engineer’s Essential Toolkit” and covered everything you could need to properly deploy, maintain, and troubleshoot a WLAN. George took extra time to emphasize how important it is to quickly get the information needed to diagnose an issue. The attendees were engaged and afterwards the conversations continued in the hallways. It’s clear that people want to know how to use the tools of the trade but not everyone has the time to learn such specific tools (I know I didn’t when I was the resident networking jack-of-all-trades).
When troubleshooting WLAN issues many factors are at play, from client drivers to ambient interference. Most times it takes nothing more than a trained eye and a packet capture to find the issue and design a plan of action. Not everyone has that trained eye though. Wi-Fi packet captures can be an overwhelming amount of information. Even seasoned veterans of Wi-Fi troubleshooting can be slowed down by trying to find that one needle in the haystack.
The engineering team here at Airtight is no stranger to this sort of thing. To make matters even harder, our support team usually assists in troubleshooting over the phone or email, meaning they never set foot in the environment or witness how clients may be (mis)behaving.
So what’s an engineer to do when they’ve got a customer with an issue and they need to quickly find it?
Build a tool to simplify, streamline, and accurately visualize wireless packet capture diagnosis. Enter Wizshark.
Wizshark was built in-house to allow our engineering teams to quickly and easily assist customers in finding issues within their WLAN. It takes packet capture (.pcap files) and injects them into a web based interface where instead of presenting a well formatted mess of data, it visualizes, tiers, and highlights important information in a simple to read graph.
I could go on for pages about the elegance and ease of use, but I think a quick video demo would convey the message a lot more clearly.
WizShark’s backend has been engineered to take full advantage of AirTight’s advanced cloud computing technologies with features such as:
- Distributed pre-processing: Once a trace is uploaded, it is divided into several chunks which are pre-processed in parallel as distinct jobs on several compute instances. The pre-processing mines information elements from the trace for visual analysis and easy filtering.
- Per-client indexing: Information elements extracted during pre-processing are cached and indexed on MAC when request is first made to view trace for that MAC. That way, when you are later analyzing that MAC’s packets, the response times are fast.
- NoSQL databases: Raw trace, pre-processed information elements and indexed information are stored in NoSQL distributed database (MongoDB). Distributed databases scale better for trace filtering queries for large data sets.
- Collaboration: Traces can be shared among users so that they can collaborate on troubleshooting.
The entire look and feel of packet analysis has been changed for the better. It’s been made easy enough for most anyone to understand and be an effective Wi-Fi troubleshooter. WizShark is now undergoing some finishing touches and is soon going to be incorporated into AirTight’s new management console, Mojo Studio.
Because good tools don’t have to be expensive and good Wi-Fi doesn’t have to be hard.
We’re offering beta access to a limited number of Wi-Fi professionals and looking for feedback before we make it part of the core suite. If you’re interested, please share this blog and video, then contact us on twitter at @airtight.
Or fill out the online form: http://go.airtightnetworks.com/wizshark.html
Also see: Network Troubleshooting in Distributed Wi-Fi Environments - tales from the field.
Don’t miss future posts! Subscribe by email.