I am seeing this from the other side. My HostHatch Singapore to Europe is fine but Singapore to Singapore is all over the place. I do not have a looking glass, just mtr from my home connection and the VPS. The cable cuts always seem to hit the same corridor. SEA-ME-WE 5 was supposed to help but we are still waiting for it here.
Another submarine cable cut near Egypt
walt22 said:
Not every provider can afford to swing traffic around Africa
Some of us remember when around Africa meant around the Cape of Good Hope via satellite. The young ones do not know how good they have it. My Leaseweb London box is steady as ever. The fiber is thinner in my memories and the patience was thicker, as I said.
SPARCstation 20, still serving HTTP
To be precise, the concentration risk I mentioned is not theoretical. Three incidents in eighteen months in the same corridor, and still no meaningful diversification by the major carriers. From Berlin, my Time4VPS node routes through Vilnius and then eastward. Vilnius to Vilnius is stable. The issue begins when Vilnius tries to reach Asia. I have documented the failover timestamps. For compliance, not for drama.
Neuland. Aber schnell.
My clients are asking about this and I want to assure everyone that our premium network at Hetzner is fully redundant with premium routing around any cable issues, please open a support ticket if you need premium assistance with your premium service, our premium support team is standing by to help my clients with any premium concerns about this premium situation, we take premium uptime very seriously
dudegoat said:
Premium network at Hetzner
We get it, dudegoat. You said this exact thing two hours ago. The word "premium" does not become more convincing with repetition.
If anyone has actual MTR from Hetzner Falkenstein or Nuremberg to Asia right now, I would like to see it. Their Singapore DC exists but European egress is what matters for this thread.
sudo make me a sandwich
LichunLars
Member
My HostHatch Stockholm is fine too, as quinnbizz noted. 12ms to Frankfurt, 180ms to everywhere I actually need, story of my life. The Stockholm to Tokyo route goes the other way entirely so we are insulated from this particular mess. Small comfort when I need to reach Singapore.
1ms or I don't want it
My $3 VPS is on HostHatch Tokyo — https://hosthatch.com — and my Minecraft server players are complaining about lag. I ran mtr to my home connection in Osaka and it is not pretty. 200ms+ where there should be 50. The server is cheap but the players are loud. Anyone know if this will actually be fixed in 48 hours or should I look for a temporary host?
instant noodles, instant deploys
ronwit said:
200ms+ where there should be 50
That is not the cable, that is your resolver timing out. Or it could be both, but check your DNS first. The dns is fine. The path to dns is not. I said this earlier and I will keep saying it until people listen.
admin said:
If anyone has actual MTR from Hetzner Falkenstein
I have Falkenstein to 8.8.8.8 and it is normal. 8.8.8.8 to anycast is not the same as Falkenstein to Singapore though.
Thanks quinnbizz. I was asking because I have a HostHatch Amsterdam box and was not seeing the same spikes, but I am only pinging Europe and US east. My Asia monitoring is nonexistent. Good to know the eastern Med path is the variable.
I have been following this thread and the one thing nobody has mentioned is price. HostHatch at $3-60 and these guys are rerouting around submarine cables. Contabo at EUR 4-60 and they are probably sweating in Nuremberg right now trying to balance their Mumbai and Singapore paths. Meanwhile Leaseweb at EUR 30-400 has the margin to absorb this. You get what you pay for, except when you do not.