Background: Integrating SSH to do traceroute network diagnostics between hosts is tedious and can be daunting if the number of hosts involved is quite large. To perform traceroute, the statistics collected must be bi-directional to be eligible for a successful network diagnostic assessment. Moreover, MTR is a command-line application, and hence, it does not display the statistics visually.
A periodic collection over time using a manual trigger method is also not recommended. One can argue to script MTR network statistics collection and execute it using Cron or Timer, but the subsequent data storage and management is another manual process. If any of these steps are not done correctly, data and other helpful information may be lost as a result.
Goals: To generate at-will network statistics reports visually with a point-and-click between hosts is also a challenging process. It is very easy to make errors, especially if you only have the statistics report in text format. The data should be presented easily so that the network administrator can focus on network solving and not fiddling the network statistics in text format. That's why our goal was to create the jwms-mtr-plugin to allow server administrators to visualize network traceroute tools output in Jenkins.
Solution & Results: Wouldn't it be nice if the network diagnostic statistics collection is done automatically? Wouldn't it be so cool if statistics could be visually presented where you simply click on the direction and understand the link statistic? jwms-mtr-plugin allows you to do EXACTLY that!
+1 Jenkins to make my life easy. From CI to CD and now to CM (continuous monitoring), Jenkins allows developers to turn their dream tool into a reality.
To do this, you need to specify the hosts you are interested in getting network diagnostic statistics from. Then let Jenkins do the remaining tedious chores for you -- from MTR running between hosts, data collections, and visual displays. Even better, it does it without an error and over an extended period so that you get a history of network diagnostic trends.
It is also possible to manually trigger the network collection, but unlike manual triggers between hosts, this solution makes it possible only to need to click a "Jenkins build now" button! It is a predictable and error-free process. Isn't this simple?!
To create this, we used: