All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
HE.net Routing Issues
Anyone else seeing HE.net routing issues?
Using a test server we have at Choopa, I cannot reach our network in HK via HE.net. No problem via NTT, but HE.net seems to be the preferred route, but traffic isn't even reaching HK, let alone our network.
1 vl217-c4-7-b2-1.pnj1.choopa.net (64.237.56.1) 1.102 ms 0.859 ms 1.262 ms 2 vl115-br1.pnj1.choopa.net (108.61.80.125) 11.772 ms vl175-br2.pnj1.choopa.net (108.61.80.129) 0.548 ms vl115-br1.pnj1.choopa.net (108.61.80.125) 11.653 ms 3 vl42-er1-q8.pnj1.choopa.net (108.61.2.90) 1.344 ms 10.64.0.26 (10.64.0.26) 1.280 ms vl41-er1-q8.pnj1.choopa.net (108.61.2.78) 1.284 ms 4 10.64.0.253 (10.64.0.253) 2.338 ms 2.299 ms 2.247 ms 5 10gigabitethernet1-2.core1.nyc6.he.net (198.32.160.61) 2.674 ms 100ge13-1.core1.nyc4.he.net (184.105.64.177) 1.446 ms 10gigabitethernet1-2.core1.nyc6.he.net (198.32.160.61) 2.559 ms 6 100ge9-1.core2.chi1.he.net (184.105.223.161) 18.232 ms 100ge13-1.core1.nyc4.he.net (184.105.64.177) 1.541 ms 100ge9-1.core2.chi1.he.net (184.105.223.161) 18.027 ms 7 100ge9-1.core2.chi1.he.net (184.105.223.161) 18.066 ms 18.007 ms 100ge14-2.core1.msp1.he.net (184.105.223.178) 28.246 ms 8 100ge14-2.core1.msp1.he.net (184.105.223.178) 28.202 ms 25.835 ms 100ge10-1.core1.sea1.he.net (184.105.223.193) 59.359 ms 9 100ge10-1.core1.sea1.he.net (184.105.223.193) 59.239 ms 59.419 ms * 10 * * * 11 * * * 12 * * * 13 * * * 14 * * * 15 * * * 16 * * * 17 * * * 18 * * * 19 * * * 20 * * * 21 * * * 22 * * * 23 * * * 24 * * * 25 * * * 26 * * * 27 * * * 28 * * * 29 * * * 30 * * *
This is quite an annoying problem with HE.net. They seem to have partial outages, but still advertise your routes, which then result in traffic not reaching you, even though you're multi-homed.
I can't be the only one seeing this.
Comments
Have you reported to the NOC? What did they say?
Yes. Still waiting for a reply.
I think it looks major enough to give them a call and start shouting.
I know you don’t have many options at the location, but i’d do everything to get rid of them for this very reason.
Yes got some HE.net related blips about 1 hour ago in Europe and Japan. Seems back to normal right now.
In case you are wondering how to format a traceroute nicely, the tag is < pre >
How does that work?
Yes, a few years ago we dropped HE.net for this very reason.
The problem is that when they go down, anyone who routes via HE.net will not be able to reach us, even if our other routes are all fine. Which seems rather insane.
Problem is, replacing HE.net with another upstream of equal capacity would probably mean spending an extra US$10K /month. Well.. maybe Cogent would be feasible.
Finally got a reply.
Apparently maintenance in Tokyo. So far 12+ hours of maintenance! No notifications of maintenance and no re-routing traffic via alternative routes... Crazy.
Mess.
Try to get higher volume. It’s possible to get good pricing on these tier 1s, comparable to HE, but only if you have volume and are a persistent negotiator ;-)
Presumably by volume you mean multiple 10G?
Cogent, frankly, is crap. Almost as crap as HE.net. Pricing is kind of similar at or above 10G. These 2 are the only providers that can do pricing comparable to the US (sort of). As far as I know, they are the only ones doing sub $1 /Mbit pricing at volume. But no one wants to use them really. They are only good for US/EU and handful of Asian traffic. So buying more would not be worth while, and going with anyone else means increasing the price by 5 - 100x more.
Looks like they resolved the issue.
Just checked how long the outage was for. At least 17 hours, given the first tickets we received was 17 hours ago.
Can do something like this:
in this case its better contact their NOC, their NOC usually answer fastly
Really ?
How is that post count for provider tag going ?