Big latency change.

[Raist]

The changes happening along your route are far more likely to be the result of changes in or a result of your ISP's established policies or those of their routing partner(s). They are the ones establishing the parameters for how you get routed to Ormuco if connecting to Canada or to asianet/JPNIC if connecting to Japan. Your ISP sets the policy that determines which routing partner to use based on analytics run for the destination ISP, and the route used to get to that partner. Then that partner makes routing decisions based on their analytics. While either of those phases may be influenced by peering/transit agreements, those would be agreements made with the ISP's, and as such are by design influenced by their policies before any content provider's needs/wants should influence routing policy.

So, basically...it points back to the ISP's mostly, except under very limited circumstances. Things like malformed data packets or something like not including a proper flag in the headers that results in shaping rules of an ISP delaying/dropping packets due to congestion. Which can still point back to that ISP because they have an issue with over utilization in general that they need to address outside of anything that SE may or may not be doing with their data.