17.3.1 Release Notes: What’s New?

cPacket has been busy working on the latest features and improvements for our 17.3.1 release and we’re excited to share these updates with you. But before we dive into what users can expect from this release, let’s begin by providing some context.

What is predictive analytics using baselines and why should we care?

The nature of today’s modern and dynamic IT environment requires that we use vast amounts of data to make sense of what’s happening around us. Predictive analytics attempts to forecast future performance issues by analyzing current and historical data. Predictive analytics has been around for years, but more and more enterprises worldwide are using it to make optimize their networks and manage them more effectively. Take for example the financial industry. With large amounts of data available and money at stake, it’s no surprise that the financial industry uses analytics to predict, detect, and reduce fraud activity.

A baseline, on the other hand, can serve several purposes and objectives. One important objective is to determine the status of the network and compare that status to standard performance guidelines. In addition, a baseline can set thresholds to alert us when the status exceeds those guidelines. When baselining is applied to network performance metrics such as bandwidth and latency, it becomes a powerful tool in understanding network behavior and can help us take a proactive versus reactive approach to troubleshooting network issues.

What’s new and improved with 17.3.1?

 1. Predictive Analytics using Baselines:

cPacket has added predictive analytics using baselines to its 17.3.1 release. This release applies to cPacket’s cStor, cVu NG series, and cClear. As seen below in Figure 1, the green solid line indicates the Actual Network Throughput through Port 20. The orange dotted line indicates the Baseline Average Throughput through Port 20. The expected throughput is based on the actual throughput that occurred in the past. Users can also go to the timeframe window to select specific dates. To obtain additional data of a specific day, simply scroll over the graph and select the day you wish to view as seen in the gray box below.

 2. Timestamp extraction from Arista switches

To analyze network event correlation and performance metrics, it’s essential to have accurate packet timestamping. Arista Networks’ 7150 series enables timestamping and because cPacket is vendor agnostic, our solutions can be deployed in any network, including Arista’s. The 7150 series supports two timestamping modes: Frame Check Sequence (FCS) replace mode and the FCS append mode. cPacket’s 17.3.1 release supports timestamp extraction for both modes.

When fast isn’t fast enough

It’s no surprise that high-speed networks place technical requirements on timestamping that are not easy to meet. Packets are transferred so quickly that microsecond resolution is not enough. As a result, many organizations have adopted a nanosecond resolution to ensure optimal network performance and accurate analysis of their data. cPacket addresses this need by providing nanosecond MiFID II compliant timestamps to provide the ability to access data and process it accurately in real time.

 3. cClear replaces SPIFEE

Lastly, as of the 17.3.1 release, cPacket’s renames SPIFEE to cClear.

To learn more about cPacket’s solutions for network performance monitoring, click here.