Update on the ASC and Indigo West cable outages

Since early September, traffic from Australia to Singapore and the rest of Southeast Asia has been going the long way round. Both subsea cables that run directly from Perth to Singapore are broken, and they won't be repaired until late October at the earliest.

Your Neptune service itself is fine. Browsing, streaming and anything hosted in Australia are unaffected. What you will notice is higher ping to servers in Singapore and nearby, which matters most if you play on SEA game servers or work with systems hosted there.

What you'll notice

On a normal day, Perth to Singapore takes around 45–50ms. With both direct cables out, traffic takes the long way: west to the Middle East, north through Guam and Japan, or across the Pacific to the US and back. Depending on the route, that can mean anywhere from 150ms to 400ms. Downloads still work and pages still load, but anything real-time to the region feels sluggish.

For Neptune customers, our transit partner GSL is currently sending Perth traffic west to Muscat in Oman and then on to Singapore. From the east coast it goes via Guam and Japan instead. If you run a traceroute to Singapore and see hops through Muscat, Guam or Tokyo, that's expected right now. Nothing is wrong with your connection. It's the detour.

Here's an mtr from our Perth PoP to a test server in Singapore this afternoon, 10 probes per hop. Latency jumps to 97ms by the time traffic reaches Muscat, and it lands in Singapore at about 166ms:

Hops 9 to 11 are routers inside the destination network that don't answer probes. That's normal and isn't packet loss: the final hop replies fine.

What's broken

There are two direct cables between Perth and Singapore, and both are down.

Indigo West has been out since 8 August, when it took a fault in the cable protection zone off Perth. That fault was repaired on 15 September, but traffic didn't come back: a second fault at the Singapore end had turned into a full break, so the cable is still down end to end.

ASC, the Australia–Singapore Cable run by Vocus, broke on 4 September in Indonesian waters, about 615 km from Singapore.

When it will be fixed

Indigo West goes first. Repairing ASC means switching off part of the system that is carrying some traffic right now, so Vocus wants a second path back before it does that. The Indigo West window has already slipped once, from around 24 October, so we expect the ASC date to move as well.

Subsea repairs depend on Indonesian permits, the repair ship's schedule, the weather, and what the crew finds when the cable comes up. We'll post an update when either cable is back.

What you can do in the meantime

Vocus still has a working path from Perth to Singapore. It runs over the working part of ASC to Indonesia, then overland through Jakarta. It's only a few milliseconds slower than the direct cable, and a long way shorter than the detour through Oman.

Our normal upstream doesn't use that path. The only way we can put your traffic on it is through a VPN instance we've set up inside Vocus's network in Perth. It's in the portal as VOCUSIX-PER1, under Australia.

Here's the same test from Perth to the same server in Singapore, this time from the VOCUSIX-PER1 instance. It gets there in about 53ms, against about 166ms on the detour:

Hop 2's numbers jump around because that router is slow to answer probes addressed to itself. Every hop after it holds steady at 53ms, which is what your traffic actually sees.

Should you use it?

If you're in Perth or elsewhere in WA and latency to Southeast Asia matters to you, it's worth a try. If you're on the east coast, your traffic has to cross the country to Perth first, so you'll see less of a gain. Try it and compare with what you get now.

If the slower route isn't bothering you, there's nothing you need to do. Everything will go back to normal by itself once the cables are repaired.

How to switch

In the portal, open VPN and either create a new tunnel or change the region on an existing one to VOCUSIX-PER1. Changing region issues new keys, so import the new config (or scan the new QR code) on your device afterwards.

Things to know

• Bandwidth is limited. It's there for gaming and other things that need low latency. Please keep bulk downloads, large backups and other heavy transfers on your usual region, so the shared capacity stays usable for everyone. Our Fair Use Policy applies here as it does to every Neptune service.
• No torrenting. As with the rest of our network, using the service to share or download content you don't have the rights to isn't allowed under our Terms and Fair Use Policy. BitTorrent traffic is blocked on this instance.
• IPv4 only. Atmosphere protective DNS isn't available on it either, so tunnels use public DNS.
• The Jakarta leg can be patchy. Vocus has reported some instability on third-party equipment in Jakarta. If your connection gets rough, switch back to your usual region.

We're watching load and latency on the instance and will make changes if it gets busy.

Once the cables are back

When either cable is repaired, traffic to Southeast Asia goes back to the direct route without you doing anything. You can then move your tunnel back to your usual region. If you'd like a hand with any of this, get in touch with our team.

Sources: Vocus repair updates, Aus Internet network status (Indigo consortium notices), Light Reading.