Titan Network "kinda" outage (not us!)

Hey all, just a quick heads-up.

Right now, I can't access the Titan Network. While that's kind of sucky for me, I've found that it may be a lot of other people affected, too. After troubleshooting what is going on, I've figured out that there is an intermediary network provider between Comcast and Linode (where the server is hosted) that has a link down. The intermediary is nLayer, and specifically, the hop between 68.31.94.118 and 69.31.95.130 is down, with very brief (a few seconds per pop) intermittent uptimes.

Just to be clear, this is not a Titan Network outage. The server is up and running, and if you happen to be on a network that doesn't go through nLayer, several of our devs have verified that you can get to all of your stuff fine. However, with these upstream provider outages, it can get hairy; it's hard to tell exactly how widespread it may be. It may just be Comcast in the Atlanta area, or it could be a regional or even a national outage across one or more networks.

If you can't get to the Titan Network, I hate to say it, but we're kind of at the mercy of nLayer. Hopefully they'll fix the connection soon.

I haven't done any extensive browsing on Titan Network, but all the pages that I've been on have been working fine for me.

I just checked my constant ping, and it seems to be consistently up now and has been for a while.

Quote:

Originally Posted by Hyperstrike This is likely local just to you Tony. Remember, when trouble spots arise, the packets are supposed to route around the trouble. The fact that you're not routing around it reliably yet means either the trouble is VERY recent or you have sub-optimal (or static) routing going on someplace where it shouldn't.

I know that's the way it's supposed to be, but working for a company with dozens of WAN links both nationally (US) and internationally, I've seen many, many cases of disconnectivity like this, even several cases of national outages when big provider network peering nodes are impacted.

Once I found the devices that were misbehaving, I didn't spend a lot of time researching nLayer's role with either Comcast or Net @ccess (Linode's upstream provider for their New Jersey datacenter), so I couldn't nail down a scope of the outage. I'm 100% certain it wasn't just me, though, which is why I posted the message. As far as I knew, it could have been a thousand people or a few million people that could have been affected.

If you do a traceroute (tracert on Windows) and go through devices on nlayer.net's network (69.31.stuff), you would have been affected. If not, you were clear. Right now, since the link is back up, everyone should be clear. I can get through to the sites fine now, and I'm still going through the 69.31.85.173 <-> 69.131.94.118 connection, which means I'm not being routed around it; they must have fixed whatever was down.