Member OP
I may be wrong, but I believe KnownHost's entire AS39201xx stopped announcing around 06:00 JST today. I noticed when my monitoring from Singapore and Tokyo both flagged unreachable. Checked a few looking glasses — no routes at all, not even partial.
I waited 30 minutes, then opened a ticket. Status page still green. No social media posts.
At 09:15 JST, routes reappeared. Latency to my instance jumped from 45ms to 120ms, then recovered to 50ms over 10 minutes. Traceroute now shows different upstream than before.
I am curious if anyone else observed this, or has insight into what upstream change would cause a complete ASN withdrawal rather than route flapping. I do not run BGP myself at this scale.
My instances in APAC were unaffected. Only EU.