Cisco ThousandEyes

My main use case for ThousandEyes  is to monitor third-party applications; we monitor BGP paths to the cloud, so if applications are not hosted in our own private cloud, we monitor what's going on in terms of third-party applications.

A specific example of a third-party application I monitor with ThousandEyes  is a clinical system called Rio; we monitor them to know the performance because they are not hosted in our own private cloud.

Although we do monitor our own applications hosted by ourselves to see what could be going on from multiple sites into our data center if something is going wrong, my main use case for ThousandEyes remains focused on third-party applications.

The best features ThousandEyes offers include monitoring page load times, assessing how long it takes for an application to load, checking for packet loss and jitter, and monitoring the routing path from the user to the server hosted in the cloud or on-premises. Additionally, you could know if the page is currently loading or not; there are still many more features, but those are what we use at the moment.

ThousandEyes has positively impacted our organization by enabling us to know if a third-party hosted application isn't functioning properly, allowing us to determine what's actually going on and whether the issue lies on our side or with the application provider.

An area where ThousandEyes can be improved is in providing more in-depth packet analysis; we've found instances where ThousandEyes indicates everything is okay, but it's actually not. I would recommend incorporating something similar to Wireshark  into ThousandEyes for deeper packet analysis as there are still aspects that ThousandEyes isn't able to determine. When something goes wrong, despite all the investments, certain results remain elusive, so they need to explore areas where Wireshark  excels.

The knowledge base around ThousandEyes is quite limited; many engineers are unaware of its existence or the benefits it can provide, and this needs to be addressed. On a rating of 1 to 10, I would give ThousandEyes a maximum of five, indicating that there are many areas for improvement, especially in helping engineers understand how to resolve issues related to third-party hosted applications, as opposed to more recognized brands such as Cisco which have extensive resources for independent study.

Lastly, the servers sometimes require rebooting to monitor applications accurately, which may indicate a potential bug or issue.

I have been using ThousandEyes for almost two years.

ThousandEyes is not very stable; sometimes you have to reboot the servers to get actual results.

Scalability with ThousandEyes is straightforward as you don't really need to scale; it's designed to monitor multiple applications, accommodating 50 or 100 applications simultaneously, and if you need multiple testing points, you just deploy agents at different parts of your network or different sites that you're managing, which it does well.

I have not previously used a different solution, so I don't have any comparisons to provide.

Before choosing ThousandEyes, I did not evaluate other options as I do not make that decision.

There have been multiple instances when the insight from ThousandEyes has helped my team solve a problem faster or avoid downtime. Before we had ThousandEyes, we were not so sure where the problem was coming from, so for example, if a third-party application goes down or becomes very slow, we could never tell whether the problem was coming from the infrastructure side or from the application provider.

The application service provider would often tell us that everything is fine on their end, but after checking our infrastructure, we discovered that every other application was fine except this one, prompting a need for in-depth analysis into what exactly is happening and where the problem is coming from so that it can be properly fixed. We then begin to track what's going on with network path visualization, checking the page load time, jitter, packet loss, and every other component.

My advice for others looking into using ThousandEyes would be to try out a demo to see if that meets their needs and solves their problems.

On a scale of 1-10, this solution rates as a 5.

We use ThousandEyes  for monitoring and troubleshooting, especially when end-users complain about latency or poor performance. We rely on ThousandEyes  to provide metrics to see if the problem is related to the local Wi-Fi or the Internet. The main goal for ThousandEyes at our company is monitoring. We also have some integration with our other monitoring systems, and we use ThousandEyes to dive deeper into specific problems when they occur.

From an engineering perspective, we achieve an excellent return on investment by fixing problems rapidly and avoiding business downtime. ThousandEyes helps reduce troubleshooting time by at least fifty percent, allowing our team to quickly support end-users and maintain business operations.

The most valuable features of ThousandEyes are the full monitoring capabilities and the ability to identify the exact location of problems in the path between the application and the end-user. For example, ThousandEyes can notify us about specific paths on the internet that are facing issues, providing detailed insights to our service providers and allowing us to raise tickets and resolve problems. The detailed information ThousandEyes provides about performance issues is crucial for us, enabling us to navigate through various network segments effectively.

When I first started using ThousandEyes, I found it challenging to extract useful information due to the numerous options available. I needed some training to effectively troubleshoot. More detailed documentation explaining the metrics and the testing mechanisms would be helpful. The information is there, but understanding exactly what ThousandEyes is testing to make its conclusions can be obscure.

We have been using ThousandEyes for about six months.

Not too much. Since we deployed ThousandEyes, it has been running pretty well.

ThousandEyes is highly scalable. The only potential issue with scaling up is related to licensing costs, as deploying more agents around the globe could become expensive. However, the product itself is good at scaling because adding more agents enhances our monitoring capabilities.

I have not interacted frequently with their support team, but based on limited experience, I would rate their customer service as very nice. I'd give it an eight out of ten.

Before using ThousandEyes, we used SolarWinds and also relied on Cisco DNA Center  for monitoring. SolarWinds did not help much with comprehensive troubleshooting, and it required ad hoc testing.

I was not part of the deployment team. From what I know, the initial setup wasn’t overly complex, and our engineers received help from Cisco to deploy the product.

Our deployment involved just one or two network engineers along with some help from Cisco, likely around one or two people on the Cisco side as well.

We achieve a very good return on investment by fixing problems quickly, thus preventing users and the business from losing time and money due to poor performance or downtime. For an engineer, the tool speeds up problem identification significantly and improves overall network performance.

The licensing cost could be a potential problem depending on the agreement, especially for smaller businesses. In our company, this was manageable due to existing agreements with Cisco. ThousandEyes licenses are based on the number of flows processed, not just the number of agents deployed.

We did not evaluate any other tools before choosing ThousandEyes.

I would highly recommend ThousandEyes because it can reduce the time spent identifying issues by at least half and provide extensive visibility into network quality.

I'd rate the solution nine out of ten.