Interest in cloud service monitoring utilities will grow as businesses become more dependent on these types of virtual services.  In this post, I’ll outline some great ways to monitor cloud services by leveraging in house hardware namely, Cisco Routers and the use of Flexible NetFlow Performance Monitoring or Medianet exports.

Performance Monitoring Flexible NetFlow (FnF) is being touted by Cisco as the technology to be used for monitoring the quality of media rich applications.  These new FnF exports provide details on TCP round trip time and for VoIP it provides jitter and packet loss metrics. 

In the first example, lets use an instance where an employee is leveraging the cloud service like Vonage.  Below, we are looking at the top 10 of 23 pages of calls.

Above you can see that most of the calls have low jitter values and notice that you can sort on packet loss by clicking on the column header.  These FnF exports can be used just as easily for monitoring video performance.   Performance Monitoring NetFlow elements allow us to confirm that a cloud service like Vonage (or even Skype) at least most of the time is working for this remote sales person and we can see that they are actively making calls and even the call duration.  

In this second example, I filtered for the cloud service salesforce.com and you can see below that scrutinizer is reporting some unhappy connection times.

Notice above that all of the different Cisco Performance Reporting options.  There are a bunch of report options that allow you to report on ranges of IP / subnets which help you get a big picture during your regular network traffic management routines.  Thresholds can be set on any combination of the above and it is all done with existing routers running IOS 15.X or more recent.    What’s great about using Flexible NetFlow is that they aren’t limited to monitoring only popular apps like Microsoft Azure, Google App Engine, etc.  This cloud monitoring solution can operate at the transport layer (e.g. TCP, UDP, etc.) or even higher layer by leveraging technologies such as Cisco NBAR which can also deliver details on RTT, availability and packet loss.

Not only can you monitor the quality of individual connections but, you can setup synthetic transactions using IP SLA and monitor / trend the values. 

What metrics can you monitor?  How about availability, response time or even bytes transferred.  Imagine being able to setup notifications if for example an entire subnet  witnesses a quality of service above 200 millisecconds for over 10 minutes.  This is the kind of monitoring and reporting that allows you to paint a clear picture of the status of your cloud services.

Solutions capable of monitoring cloud services are still evolving and without a doubt, the metrics mentioned in this post will be part of the evolution.