KnownHost finally issued a statement. "We experienced extended connectivity issues in our Amsterdam location due to upstream provider maintenance." Upstream provider maintenance. Not a fire. Not a fire that was on the news.
That major datacenter fire—who's actually affected?
tinaworm
Member OP
jar19 said:
Not a fire. Not a fire that was on the news.
I found the news. Dutch news site NU.nl, May 7, "Brand in datacenter Amsterdam." Leaseweb confirmed to them it was their facility. KnownHost is just... not reading the news? Hoping customers do not?
worst bandwidth, best stories
I was considering Leaseweb for a project in Singapore. Now I am considering Equinix instead. Price is double but at least they have actual fire suppression, not whatever this was.
jam_act said:
Actual fire suppression
We do not know what suppression they had. Could have been Inergen, could have been water, could have been a guy with a CO2 bottle. The lack of detail in all statements is the problem.
one small ping for man...
Does anyone know if ByteMole ever issued a statement? The first post says their statement was cut off.
tinaworm
Member OP
ultrareal said:
ByteMole ever issued a statement?
I never finished that. ByteMole posted on Twitter/X: "We are aware of reports regarding a facility in Los Angeles and are investigating. More details to follow." That was May 7. Nothing since. They are a tiny reseller, probably do not even have direct presence there.
worst bandwidth, best stories
So to summarize what this thread actually established:
- CloudCone: down 4-6 hours, IP reassignments, support dead, email reputation destroyed
- Leaseweb: down 14-22 hours, metrics lost, billing unaffected, honest-ish statement
- Hostinger: 9 hour migration, status page suspicious timing
- KnownHost: 48+ hours, no real statement, "upstream maintenance" lie
- ByteMole: probably just a reseller, no follow-up
The winner for worst handling is KnownHost by a lot.
jordtier said:
Worst handling is KnownHost
I would say CloudCone for the email reputation alone. That damage lasts months. KnownHost customers can migrate. CloudCone customers with dead SPF records cannot even get into their control panels to fix the root cause. — https://www.spamhaus.org
PacketpunkUma
Member
My homelab is in my closet and the only fire risk is a cheap PSU from AliExpress. After reading this thread I am oddly comforted by this.
Did anyone ever figure out what started the fire? I saw speculation about battery storage but nothing confirmed.