Status: resolved. Between 03:11 and 03:41 UTC on 13 October, connections through our Singapore gateway were slow and some of them failed.
What caused it
The cause was an unusually large burst of traffic: about 560,000 new connections in 30 minutes, opened from around 30 servers at the same time. At that rate the gateway could not set up new routes fast enough, so new connections had to wait in line or timed out.
What we changed
The gateway recovered at 03:41 UTC, and we have since upgraded how it handles bursts like this one. You can now run that many simultaneous connections through the gateway without slowing it down for anyone else.
What to do
Nothing needs changing. If anything still looks slow through Singapore, contact our support team and we will check it right away. You can read how gateways and exits relate in the setup guide.