All North American Data Center Worlds Network Technical Difficulties (Oct. 26)

[BurnzieThornton]

It looks to me like the problem could be resolved by Ormuco, who is a fiber provider and looks to be the cloud provider for the Aether NA server cluster in Montreal.

In theory, Ormuco could solve the issue by shifting traffic away from their Level3 transit and towards another transit peer (Cogent, Verizon, or Tata America just from a quick glance).

I would assume Eidos would either directly handle this traffic engineering exercise directly with the proper communities between them and Ormuco, or have Ormuco manually implement a simple traffic engineering solution for Eidos' 199.91.189.0/24 allocation at a minimum.

Regardless, if some people can play consistently on the Aether cluster then this is a network issue that should be resolved by Eidos and their immediate upstream Ormuco.

Putting this all on a Level3 fiber cut in Houston is disingenuous, as given adequate capacity they could cut them out of the equation entirely for the duration of the issue (and 14 hours deep into an issue like this is somewhat unforgivable from an operations standpoint of operating a content network with multiple upstreams).