monitoring – Sudden peak in incoming bytes in HAProxy

I was seeing the data in real time in Google Analytics and I noticed that visits to the pages were suddenly reduced by 50% for a small period of time. The fall occurred at 6.28 p.m.

When I realized this, I went into the Datadog control panel and immediately saw the following:

enter the description of the image here

That peak happened at about the same time. The graph above is haproxy.frontend.bytes.in_rate, which Datadog defines as "Byte rate in frontend hosts".

It jumped from 18 bytes per second to 53 bytes per second.

Does anyone know what could have caused this or how could I further debug these types of peaks?

Thank you,
Wayne