I suppose it's mildly unsatisfactory as it leaves you wondering if traffic that must go via that route is somehow less performant; changing the routing for BQM is a bit of a frig (if that's what has happened).
I suppose it's mildly unsatisfactory as it leaves you wondering if traffic that must go via that route is somehow less performant; changing the routing for BQM is a bit of a frig (if that's what has happened).
It's still a bit of a mystery - we have other ping graphs to endpoints over the same LINX connection and don't see problems - we only see these latency spikes on BQM, when the traffic goes over LINX. That LINX connection k.aimless) usually peaks at about 40% at peak - which is usually evenings anyway.