Category Archives: Reliability

Rising Outages Mean Days Without the Internet Could Soon be Reality

By Paul Gampe

Most of us experience internet outages on an individual — and generally irregular — level. Perhaps we’re unable to check in with friends on social media, stream a movie or pull up a recipe for that evening’s dinner. The inconvenience is frustrating, but the impact is limited. 2016 may not have been when many people first became aware of the vulnerabilities featured in the underlying routing architecture of the public internet, but it remains a watershed year where these vulnerabilities became such prominent and repeated targets. This yielded a wider impact on communities and businesses from large-scale internet outages, caused by deliberate and malicious cyber-attacks. As industries, services and governments have grown more reliant on the public internet, malicious characters…

Is there Static in the Cloud? – Part 2

By William B. Norton

The Mysterious Inter-Cloud Connectivity Problems Last week I walked a couple cloud techs through my current white paper “Cloud Interconnections.” This white paper compares the interconnection models of AWS, GCP, and MAZ. It turns out that each cloud provider chose different names for their cloud services, and the interconnection models are different enough to be confusing. So I wrote a white paper that documented the difference between the various interconnection models, along with a mini “Business Case for Direct Connection” for each. Stay tuned- I’ll post snippets from this white paper and do a full white paper walkthrough as a one hour webinar next month. But first, as part of the process, I am walking people through the white paper privately to make sure…

Is there Static in the Cloud? – Part 1

By William B. Norton

Dr.Peering in the Cloud I wanted to compare the connectivity within the clouds to the connectivity into the clouds, so I deployed about 20 VMS across Amazon Web Services, Google Cloud Platform, Microsoft Azure and had them ping each other.  I was expecting a completely green matrix, one that showed flawless connectivity between the major cloud providers, but then something else happened (see matrix below). On the left axis are my agent regions, and across the top are the agent regions that they are pinging. It is tiny, but the numbers in each cell of the table are median latency measures. The coloring of the cells reflects deviation from the median. Red means that there have been measures today that deviate 50% from the median!…

Cloud Interconnection Models

By William B. Norton

The Internet is evolving into two hemispheres: the Public Internet and the Private Internet. The public Internet is the shared best-effort packet delivery vehicle that we all know and love. The private Internet is attached to the side of the public Internet, and consists of dedicated private (not-shared) connections between entities. The private connection is architectural; you set it and forget it. The routing algorithms automatically flow traffic along the shortest path. Why does this Public-Private Internet distinction matter? Public Internet traffic is intermingled with others Internet traffic. This has some very important implications. First, the good news. Shared resources can lead to economies of scale and aggregation efficiencies. By aggregation efficiencies I mean that the peaks in one flow may be offset…