Easy solution to measure consumer expertise? ThousandEyes on Raspberry Pis


In Cisco IT we now have loads of follow bringing acquired corporations into the fold. We’ve added fourteen from January 2020 via August 2021 alone. We’re accountable for connecting the brand new places of work to our community and delivering shared providers like Cisco Webex, Workplace 365, and Salesforce.

When Cisco acquired Acacia Communications in March 2021, we would have liked to carry their world places of work and contract producers onto our community. Earlier than changing Acacia’s SD-WAN with our personal enterprise spine we had to verify the change wouldn’t have an effect on the consumer expertise in Acacia’s places of work.

For earlier acquisitions, we used Cisco ThousandEyes to measure the consumer expertise from every location to Cisco providers like Webex. ThousandEyes is a part of our Full-Stack Observability (FSO) suite—together with Cisco AppDynamics to observe the applying expertise and Cisco Intersight to troubleshoot Cisco UCS rack servers. Our plan: Use ThousandEyes to measure latency, packet loss, and out there bandwidth between Acacia places and cloud providers to determine a benchmark. Then examine these outcomes in opposition to the identical outcomes for close by Cisco places.


The twist—we didn’t have entry to Acacia’s routers

Ordinarily we measure consumer expertise at Cisco and newly acquired corporations by putting in ThousandEyes on routers and endpoints. However on this case, we couldn’t get community entry to put in ThousandEyes as a result of Acacia was in the course of a brand new product launch. We would have liked one other solution to set up the agent.

Our answer? Set up ThousandEyes Enterprise Brokers on low-cost, credit-card sized computer systems referred to as Raspberry Pis and ship one to every Acacia location. Raspberry Pis value lower than $100 apiece, and we’d have the ability to reuse them for future acquisitions.


As straightforward as pi

Putting in and configuring ThousandEyes on every Raspberry Pi took lower than half-hour. Then we simply shipped one preconfigured system to every Acacia website. We included easy directions—primarily, “Please plug this in close to an worker’s desk and join it to the community.” As soon as related, the Raspberry Pi routinely phoned residence to our ThousandEyes server to examine for software program updates. Achieved!

Delivery {hardware} to India takes a very long time due to import restrictions, so in Acacia’s Bangalore workplace, we deployed ThousandEyes on a digital machine as a substitute.


Testing the end-to-end consumer expertise

I spent about an hour on the ThousandEyes net interface establishing artificial assessments to measure latency, packet loss, and bandwidth between every set of two Acacia places. I additionally wrote a take a look at to simulate the consumer expertise when loading the house web page for cloud providers like Cisco Webex, Workplace 365, and Salesforce.

Checks run each half-hour. ThousandEyes makes it straightforward to identify excessive latency. In Determine 1, the spike exhibits excessive latency for visitors flowing from Acacia’s Ottawa, Canada workplace to Holmdel, New Jersey.

Determine 1. Latency Site visitors

Determine 2 exhibits latency alongside the community path from 10 of Acacia’s websites to their Massachusetts knowledge heart. It isolates a connectivity problem with the Acacia Shenzhen website to a single hop alongside the trail—worthwhile info that helped our acquisition and operations groups perceive the impression of the problem and resolve it.

Determine 2. Path Validation

I additionally added a view that plots the values on a map of the world.

Determine 3. Latency values on a world map


Backside line: we perceive the consumer expertise on a community we are able to’t entry

Summing up, ThousandEyes on Raspberry Pis provides us a easy solution to examine that new workers from an acquisition will proceed to have the identical nice expertise after they transfer onto our community. If not, we’ll make modifications. Configuring the Raspberry Pis and establishing the ThousandEyes assessments took simply half a day. And all Acacia needed to do was plug in a tiny laptop to face in for a consumer laptop computer. Such a sublime answer requires a poem:

ThousandEyes on Raspberry Pis
Arrange is a breeze
Raspberry Pis for Thousand Eyes
Take a look at latency with ease


Study extra about our journey to a sophisticated community
structure by clicking via our interactive journey map

Comply with Cisco IT on social!



Study extra

ThousandEyes net web page

Workers do their very own efficiency troubleshooting with ThousandEyes

Zeroing in on community efficiency points with ThousandEyes



Please enter your comment!
Please enter your name here