Network congestion 28 Feb?

gumbuya 10 years ago updated by Konstantin (Here to help) 10 years ago 1
Yesterday night, roughly 1400hrs UTC we were alerted by multiple services with slow response times. This include services hosted in West US, Southeast Asia, and also Amazon's CloudFront CDN edge servers.

It's strange that our services in different parts of the globe can simultaneously become slow, and/or that all of Anturis' monitoring nodes (USA, Canada, Germany) face equal congestion. Throughout the night it somewhat reduced to sporadic incidents and quick subsequent recovery, but this behaviour has been popping out throughout the day even up to now.

Despite adjusting the response times and failure count levels to make it more tolerant, the underlying network seems to "adjust" itself accordingly, going beyond those levels and continuing to raise alerts to us.

Is there some real underlying network congestion happening between the monitoring nodes during this period?
Yes we also observed internet problem esp, from our German location during this period of time. We will further investigate the problem and come back to you.